IMI – Myopia Genetics Report

Abstract

The knowledge on the genetic background of refractive error and myopia has expanded dramatically in the past few years. This white paper aims to provide a concise summary of current genetic findings and defines the direction where development is needed.

We performed an extensive literature search and conducted informal discussions with key stakeholders. Specific topics reviewed included common refractive error, any and high myopia, and myopia related to syndromes.

To date, almost 200 genetic loci have been identified for refractive error and myopia, and risk variants mostly carry low risk but are highly prevalent in the general population. Several genes for secondary syndromic myopia overlap with those for common myopia. Polygenic risk scores show overrepresentation of high myopia in the higher deciles of risk. Annotated genes have a wide variety of functions, and all retinal layers appear to be sites of expression.

The current genetic findings offer a world of new molecules involved in myopiagenesis. As the missing heritability is still large, further genetic advances are needed. This Committee recommends expanding large-scale, in-depth genetic studies using complementary big data analytics, consideration of gene-environment effects by thorough measurement of environmental exposures, and focus on subgroups with extreme phenotypes and high familial occurrence. Functional characterization of associated variants is simultaneously needed to bridge the knowledge gap between sequence variance and consequence for eye growth.

1. Summary

For many years, it has been recognized that myopia is highly heritable, but only recently has significant progress been made in dissecting the genetic background. In particular genome-wide association studies (GWAS) have successfully identified many common genetic variants associated with myopia and refractive error. It is clear that the trait is complex, with many genetic variants of small effect that are expressed in all retinal layers, often with a known function in neurotransmission or extracellular matrix. Exact mechanisms by which these genes function in a retina-to-sclera signaling cascade and other potential pathways remain to be elucidated. The prediction of myopia from genetic risk scores is improving, but whether this knowledge will affect clinical practice is yet unknown. This Committee recommends expanding large-scale genetic studies to further identify the molecular mechanisms through which environmental influences cause myopia (gene-by-environment effects), with an ultimate view to develop targeted treatments.

2. Key Points

1. Refractive errors including myopia are caused by a complex interplay between many common genetic factors and environmental factors (near work, outdoor exposure).

2. Early linkage studies and candidate gene studies have identified up to 50 loci and genes, but findings remained mostly unverified in replication studies.

3. Large consortia performing GWAS enabled identification of common genetic variants associated with refractive error and myopia.

4. The Consortium for Refractive Error and Myopia (CREAM) and 23andMe published findings from GWAS separately, and later combined studies in a GWAS meta-analysis, identifying 161 common variants for refractive error but explaining only approximately 8% of the phenotypic variance of this trait.

5. Polygenic risk scores based on these variants indicate that persons at high genetic risk have an up to 40 times increased risk of myopia compared with persons at low genetic risk.

6. The genetic loci appear to play a role in synaptic transmission, cell-cell adhesion, calcium ion binding, cation channel activity, and the plasma membrane. Many are light-dependent and related to cell-cycle and growth pathways.

7. Pathway analysis confirms the hypothesis for a light-induced retina-to-sclera signaling pathway for myopia development.

8. Genome-environment-wide interaction studies (GEWIS) assessing variant × education interaction effects identified nine other loci. Evidence for statistical interaction was also found; those at profound genetic risk with higher education appeared particularly susceptible to developing myopia.

9. As most of the phenotypic variance of refractive errors is still unexplained, larger sample sizes are required with deeper coverage of the genome.

10. The ultimate aim of genetic studies is to discern the molecular signaling cascade and open up new avenues for intervention.

3. Introduction

Although myopia is strongly determined by environmental factors, the trait has long been known to run in families, suggesting a genetic predisposition. The heritability of refractive error, using spherical equivalent as a quantitative trait, has been determined in a number of families and twin studies.^1–8 The estimates resulting from these studies calculated heritabilities from 15% to 98%.^5,7–10 However, it is important to note that this does not necessarily imply that most refractive error is genetic; familial clustering also can be determined by other factors.¹¹

Like many other traits, common myopia has a complex etiology that is influenced by an interplay of genetic and environmental factors.¹² The current evidence, as summarized in this review, indicates that it is likely to be caused by many genes, each contributing a small effect to the overall myopia risk. The evidence for this has been confirmed by large GWAS.^{1–5,7,13,14} Several high, secondary syndromic forms of myopia, such as Marfan, Stickler, and Donnai-Barrow, form the exception, as they inherit predominantly in a Mendelian fashion with one single, highly penetrant, causal gene.¹⁵

This white paper aims to address the recent developments in genetic dissection of common refractive errors, in particular myopia. Up until the era of GWAS, identification of disease-associated genes relied on studies using linkage analysis in families or investigating variants in candidate genes. In myopia, these were singularly unsuccessful, and before 2009, there were no genes known for common myopia occurring in the general population. However, with the advent of GWAS, many refractive error genes associated with myopia have been identified, providing potential new insights into the molecular machinery underlying myopia, and perhaps promising leads for future therapies.

4. Heritability

Eighty years ago, Sir Duke-Elder was one of the first to recognize a “hereditary tendency to myopia.”¹⁶ Since then, evidence for familial aggregation has been delivered by various familial clustering, twin, and offspring studies,^1–4 and a genetic predisposition became more widely recognized. Strikingly, the estimates of myopia heritability vary widely among studies, with values as low as 10%^9,10 found in a parent-offspring study in Eskimos, to as high as 98% in a study of female twin pairs^5,7,8 (Table 1). Differences in study design and method of analysis may account for this, but it is also conceivable that the phenotypic variance determined by heritable factors is high in settings in which environmental triggers are limited, and low where they are abundant. Based on literature, heritability of myopia is probably between 60% and 80%.

Table 1.

Heritability Estimates of Refractive Error

Subjects	Study	Heritability Estimate (±SE or 95% CI)
Monozygous and dizygous twin pairs	Dirani et al. 20066	0.88 ± 0.02 (men) (SE)
	Hammond et al. 200121	0.86 (0.83–0.89)
	Lyhne et al. 20017	0.89–0.94 (0.82–0.96)
Sibling pair	Guggenheim et al. 2007152	0.90 (0.62–1.12)
	Peet et al. 2007153	0.69 (0.58–0.85)
Full pedigree	Klein et al. 200919	0.62 ± 0.13
Parent-offspring pair	Lim et al. 2014154	0.30 (0.27–0.33)

Variation in corneal curvature and axial length contribute to the degree of myopia.¹⁷ Twin studies also estimated a high heritability for most of the individual biometric parameters.^18,19 Correlations between corneal curvature and axial length were at least 64%,²⁰ suggesting a considerable genetic overlap between the parameters.

Studies addressing the inheritance structure of myopia and its endophenotypes identified several models, mostly a combination of additive genetic and environmental effects.^6,18,21,22 Genome-wide complex trait analysis, using high-density genome-wide single-nucleotide polymorphism (SNP) genotype information, was performed in young children from the Avon Longitudinal Study of Parents and Children (ALSPAC), and results suggested that common SNPs explained approximately 35% of the variation in refractive error between unrelated subjects.²³ SNP heritability calculated by linkage disequilibrium score regression in the CREAM Consortium was 21% in European individuals but only 5% in Asian individuals, which could be due to the low representation of this ancestry.²⁴

In conclusion, the genetic component of myopia and ocular biometry is well recognized, but its magnitude varies in studies depending on the population being studied, the study design, and the methodology. It is important to note that the recent global rise of myopia prevalence is unlikely to be due to genetic factors, but the degree of myopia may still be under genetic control.²⁵

5. Linkage Studies

A number of linkage studies for myopia were performed in families and high-risk groups before the GWAS era (Fig. 1).²⁶ Linkage studies have searched for cosegregation of genetic markers (such as cytosine-adenine [CA] repeats) with the trait through pedigrees, and has been successfully applied for many Mendelian disorders.²⁷ In families with an autosomal dominant inheritance pattern of myopia, this approach helped to identify several independent loci for (high) myopia: MYP 1 to 20,^26,28–30 as well as several other loci.^31–36 Fine-mapping of several of these loci led to candidate genes, such as the IGF1 gene located in the MYP3 locus.¹² Although validation of the same markers failed in these candidate genes, other variants appeared associated with common myopia, suggesting genetic overlap between Mendelian and complex myopia.³⁷ Linkage studies using a complex inheritance design found five additional loci.^38–42

Figure 1.

Historic overview of myopia gene finding. Genes identified using WES are marked as purple. Other loci (linkage studies, GWAS) are marked as red.

With the development of new approaches for gene finding, linkage analysis with CA-markers became unfashionable. Nevertheless, segregation and linkage analysis of a variant or region in pedigrees is still a common procedure for fine-mapping or dissection of disease haplotypes.

6. Secondary Syndromic Myopia

Myopia can accompany other systemic or ocular abnormalities. The secondary syndromic myopias are generally monogenic and have a wide spectrum of clinical presentations. Table 2 summarizes the known syndromic conditions that present with myopia, and Table 3 summarizes the known ocular conditions.⁴³ Among these disorders are many mental retardation syndromes, such as Angelman (Online Mendelian Inheritance in Man database [OMIM] #105830), Bardet-Biedl (OMIM #209900), and Cohen (OMIM #216550) and Pitt-Hopkins syndrome (OMIM #610954). Myopia also can be a characteristic feature in heritable connective tissue disorders, such as Marfan (OMIM #154700), Stickler (OMIM #108300, #604841, #614134, #614284), and Weill-Marchesani syndrome (OMIM #277600, #608328, #614819, #613195), and several types of Ehlers-Danlos syndrome (OMIM #225400, #601776).

Table 2.

Overview of Secondary Syndromic Forms of Myopia: Systemic Syndromes Associated With Myopia

Title	Gene and Inheritance Pattern
Acromelic frontonasal dysostosis	ZSWIM6 (AD)
Alagille syndrome	JAG1 (AD)
Alport syndrome	COL4A5 (XLD); COL4A3 (AR/AD)
Angelman syndrome	UBE3A (IP); CH
Bardet-Biedl syndrome	ARL6; BBS1; BBS2; BBS4; BBS5; BBS7; BBS9; BBS10; BBS12; CEP290; LZTFL1; MKKS; MKS1; SDCCAG8; TMEM67; TRIM32; TTC8; WDPCP (AR)
Beals syndrome	FBN2 (AD)
Beaulieu-Boycott-Innes syndrome	THOC6 (AR)
Bohring-Opitz syndrome	ASXL1 (AD)
Bone fragility and contractures; arterial rupture and deafness	PLOD3 (AR)
Branchiooculofacial syndrome	TFAP2A (AD)
Cardiofaciocutaneous syndrome	MAP2K2 (AD)
Cohen syndrome	VPS13B (AR)
Cornelia de Lange syndrome	NIPBL (AD); HDAC8 (XLD)
Cowden syndrome	PTEN (AD)
Cranioectodermal dysplasia	IFT122 (AR)
Cutis laxa	ATP6V0A2; ALDH18A1 (AR)
Danon disease	LAMP2 (XLD)
Deafness and myopia	SLITRK6 (AR)
Desanto-Shinawi syndrome	WAC (AD)
Desbuquois dysplasia	CANT1 (AR)
Donnai-Barrow syndrome	LRP2 (AR)
DOORS	TBC1D24 (AR)
Ehlers-Danlos syndrome	COL5A1 (AD); PLOD1 (AR); CHST14 (AR); ADAMTS2 (AR); B3GALT6 (AR); FKBP14 (AR)
Emanuel syndrome	CH
Fibrochondrogenesis	COL11A1 (AR)
Gyrate atrophy of choroid and retina with/without ornithinemia	OAT (AR)
Hamamy syndrome	IRX5 (AR)
Homocystinuria	CBS (AR)
Joint laxity; short stature; myopia	GZF1 (AR)
Kaufman oculocerebrofacial syndrome	UBE3B (AR)
Kenny-Caffey syndrome	FAM111A (AD)
Kniest dysplasia	COL2A1 (AD)
Knobloch syndrome	COL18A1 (AR)
Lamb-Shaffer syndrome	SOX5 (AD)
Lethal congenital contracture syndrome	ERBB3 (AR)
Leukodystrophy	POLR1C; POLR3A; POLR3B; GJC2 (AR)
Linear skin defects with multiple congenital anomalies	NDUFB11; COX7B (XLD)
Loeys-Dietz syndrome	TGFBR1; TGFBR2 (AD)
Macrocephaly/megalencephaly syndrome	TBC1D7 (AR)
Marfan syndrome	FBN1 (AD)
Marshall syndrome	COL11A1 (AD)
Microcephaly with/without chorioretinopathy; lymphedema; and/or mental retardation	KIF11 (AD)
Mohr-Tranebjaerg syndrome	TIMM8A (XLR)
Mucolipidosis	GNPTAG (AR)
Muscular dystrophy	TRAPPC11; POMT; POMT1; POMT2; POMGNT1; B3GALNT2; FKRP; DAG1; FKTN (AR)
Nephrotic syndrome	LAMB2 (AR)
Noonan syndrome	A2ML1; BRAF; CBL; HRAS; KRAS; MAP2K1; MAP2K2; NRAS; PTPN11; RAF1; RIT1; SOS1; SHOC2; SPRED1 (AD)
Oculocutaneous albinism	TYR (AR)
Oculodentodigital dysplasia	GJA1 (AR)
Pallister-Killian syndrome	CH
Papillorenal syndrome	PAX2 (AD)
Peters-plus syndrome	B3GLCT (AR)
Pitt-Hopkins syndrome	TCF4 (AD)
Pontocerebellar hypoplasia	CHMP1A (AR)
Poretti-Boltshauser syndrome	LAMA1 (AR)
Prader-Willi syndrome	NDN (PC); SNRPN (IP); CH
Pseudoxanthoma elasticum	ABCC6 (AR)
Renal hypomagnesemia	CLDN16; CLDN19 (AR)
SADDAN	FGFR3 (AD)
Schaaf-Yang syndrome	MAGEL2 (AD)
Schimke immunoosseous dysplasia	SMARCAL1 (AR)
Schuurs-Hoeijmakers syndrome	PACS1 (AD)
Schwartz-Jampel syndrome	HSPG2 (AR)
Sengers syndrome	AGK (AR)
Short stature; hearing loss; retinitis pigmentosa and distinctive facies	EXOSC2 (AR)
Short stature; optic nerve atrophy; and Pelger-Huet anomaly	NBAS (AR)
SHORT syndrome	PIK3R1 (AD)
Short-rib thoracic dysplasia with/without polydactyly	WDR19 (AR)
Shprintzen-Goldberg syndrome	SKI (AD)
Singleton-Merten syndrome	IFIH1 (AD)
Small vessel brain disease with/without ocular anomalies	COL4A1 (AD)
Smith-Magenis syndrome	RAI1 (AD)
Spastic paraplegia	HACE1 (AR)
Split hand/foot malformation	CH
Stickler syndrome	COL2A1 (AD); COL11A1 (AD); COL9A1 (AR); COL9A2 (AR)
Syndromic mental retardation	SETD5 (AD); MBD5 (AD); USP9X (XLD); NONO (XLR); RPL10 (XLR); SMS (XLR); ELOVL4 (AR); KDM5C (XLR)
Syndromic microphthalmia	OTX2; BMP4 (AD)
Temtamy syndrome	C12orf57 (AR)
White-Sutton syndrome	POGZ (AD)
Zimmermann-Laband syndrome	KCNH1 (AD)

AD, autosomal dominant; AR, autosomal recessive; CH, chromosomal; IP, imprinting defect; XLD, × linked dominant; XLR, × linked recessive.

Table 3.

Overview of Secondary Syndromic Forms of Myopia: Ocular Syndromes Associated With Myopia

Title	Gene and Inheritance Pattern
Achromatopsia	CNGB3 (AR)
Aland Island eye disease	GPR143 (XLR)
Anterior-segment dysgenesis	PITX3 (AD)
Bietti crystalline corneoretinal dystrophy	CYP4V2 (AD)
Blue cone monochromacy	OPN1LW; OPN1MW (XLR)
Brittle cornea syndrome	ZNF469; PRDM5 (AR)
Cataract	BFSP2; CRYBA2; EPHA2 (AD)
Colobomatous macrophthalmia with microcornea	CH
Cone dystrophy	KCNV2 (AD)
Cone rod dystrophy	C8orf37 (AR); RAB28 (AR); RPGR (XLR); CACNA1F (XLR)
Congenital microcoria	CH
Congenital stationary night blindness	NYX (XLR); CACNA1F (XLR); GRM6 (AR); SLC24A1 (AR); LRIT3 (AR); GNB3 (AR); GPR179 (AR)
Ectopia lentis et pupillae	ADAMTSL4 (AR)
High myopia with cataract and vitreoretinal degeneration	P3H2 (AR)
Keratoconus	VSX1 (AD)
Leber congenital amaurosis	TULP1 (AR)
Microcornea, myopic chorioretinal atrophy, and telecanthus	ADAMTS18 (AR)
Microspherophakia and/or megalocornea, with ectopia lentis and/or secondary glaucoma	LTBP2 (AR)
Ocular albinism	OCA2 (AR)
Primary open angle glaucoma	MYOC; OPTN (AD)
Retinal cone dystrophy	KCNV2 (AR)
Retinal dystrophy	C21orf2 (AR); TUB (AR)
Retinitis pigmentosa	RP1 (AD); RP2 (XLR); RPGR (XLR); TTC8 (AR)
Sveinsson chorioretinal atrophy	TEAD1 (AD)
Vitreoretinopathy	ZNF408 (AD)
Wagner vitreoretinopathy	VCAN (AD)
Weill-Marchesani syndrome	ADAMTS10 (AR); FBN1 (AD); LTBP2 (AR); ADAMTS17 (AR)

A number of inherited retinal dystrophies also present with myopia, most strikingly X-linked retinitis pigmentosa caused by mutations in the RPGR-gene (retinal G protein–coupled receptor) (see Ref. 44 for common gene acronyms) and congenital stationary night blindness.⁴⁵ Other eye disorders accompanied by myopia are ocular albinism (OMIM #300500) and Wagner vitreoretinopathy (OMIM #143200).

Most genes causing syndromic forms of myopia have not (yet) been implicated in common forms of myopia, except for collagen type II alpha 1 chain (COL2A1)^46,47 and fibrilin 1 (FBN1).^24,48 However, a recent study screened polymorphisms located in and around genes known to cause rare syndromic myopia, and found them to be overrepresented in GWASs on refractive error and myopia.⁴⁹ This implies that although rare, pathogenic mutations in these genes have a profound impact on the eye; more benign polymorphisms may have only subtle effects on ocular biometry and refractive error.

7. Candidate Gene Studies

Candidate genes are generally selected based on their known biological, physiological, or functional relevance to the disease. Although sometimes highly effective, this approach is limited by its reliance on existing knowledge. Another caveat not specific for this approach is that genetic variability across populations can make it difficult to distinguish normal variation from disease-associated variation.¹³ In addition, candidate gene studies are very prone to publication bias, and therefore published results are highly selected.

Numerous genes have been investigated in candidate gene studies for refractive error traits. Table 4 summarizes all studies that reported statistically significant associations for myopia or ocular refraction. Genes that encode collagens (COL1A1, COL2A1),^46,47 transforming growth factors (TGFβ1, TGFβ2, TGFβ-induced factor homeobox 1 [TGIF1]),^50–52 hepatocyte growth factor and its receptor (HGF, CMET),^53–55 insulin-like growth factor (IGF1),^56,57 matrix metalloproteinases (MMP1, MMP2, MMP3, MMP9, MMP10),^58,59 the lumican gene (LUM),⁶⁰ and the ocular developmental gene PAX6,⁶¹ all showed promise in candidate gene studies. Unfortunately, like myopia linkage studies, these studies generally lacked validation by independent studies.⁶² Meta-analyses combining data from several candidate gene studies provided evidence for a consistent association between a single SNP in the PAX6 gene and extreme and high myopia.⁶³ Meta-analyses of the LUM and IGF1 genes did not confirm an association.^64,65

Table 4.

Summary of Candidate Gene Studies Reporting Positive Association Results With Myopia

Gene	Study	Ethnicity	Independent Confirmation	Replication in GWAS
APLP2	Tkatchenko et al. 2015131	Caucasian	–	–
BMP2K	Liu et al. 2009155	Chinese	–	–
CHRM1	Lin et al. 2009156	Han Chinese	X157	–
CHRM1	Guggenheim et al. 2010158	Caucasian	X157	–
CMET	Khor et al. 200955	Chinese	–	–
COL1A1	Inamori et al. 2007159	Japanese	–	–
COL2A1	Mutti et al. 200746	Caucasian	–	–
COL2A1	Metlapally et al. 200947	Caucasian	–	–
CRYBA4	Ho et al. 2012160	Chinese	–	–
HGF	Han et al. 200654	Han Chinese	–	–
HGF	Yanovitch et al. 2009161	Caucasian	–	–
HGF	Veerappan et al. 201053	Caucasian	–	–
IGF1	Metlapally et al. 201057	Caucasian	–	–
LUM	Wang et al. 200660	Chinese	–	–
LUM	Chen et al. 2009162	Han Chinese	–	–
LUM	Lin et al. 2010164	Chinese	–	–
LUM	Guggenheim et al. 2010158	Caucasian	–	–
MFN1	Andrew et al. 2008164	Caucasian	X165	–
MMP1	Wojciechowski et al. 2010130	Amish	–	–
MMP1	Wojciechowski et al. 201359	Caucasian	–	–
MMP10	Wojciechowski et al. 201359	Caucasian	–	–
MMP2	Wojciechowski et al. 2010130	Amish	–	–
MMP2	Wojciechowski et al. 201359	Caucasian	–	–
MMP3	Hall et al. 200958	Caucasian	–	–
MMP9	Hall et al. 200958	Caucasian	–	–
MYOC	Tang et al. 200763	Chinese	–	–
MYOC	Vatavuk et al. 2009167	Caucasian	–	–
MYOC	Zayats et al. 2009168	Caucasian	–	–
PAX6	Tsai et al. 2008169	Chinese	–	–
PAX6	Ng et al. 2009170	Han Chinese	–	–
PAX6	Han et al. 2009171	Han Chinese	–	–
PAX6	Miyake et al. 2012172	Japanese	–	–
PAX6	Kanemaki et al. 2015173	Japanese	–	–
PSARL	Andrew et al. 2008164	Caucasian	–	–
SOX2T	Andrew et al. 2008164	Caucasian	–
TGFβ1	Lin et al. 200650	Chinese	–	X24
TGFβ1	Zha et al. 2009174	Chinese	–	X24
TGFβ1	Khor et al. 201056	Chinese	–	X24
TGFβ1	Rasool et al. 2013175	Indian	–	X24
TGFβ2	Lin et al. 200951	Han Chinese	–	–
TGIF	Lam et al. 200352	Chinese	–	–
TGIF1	Ahmed et al. 201452,176	Indian	–	–
LAMA1	Zhao et al. 2011177	Chinese	–	–
UMODL1	Nishizaki et al. 2009178	Japanese	–	–

X indicates independent conformation or replication in GWAS study with reference included.

8. Genome-Wide Association Studies

Since the first GWAS in 2005,⁶⁶ more than 3000 human GWAS have examined more than 1800 diseases and traits, and thousands of SNP associations have been found. This has greatly augmented our knowledge of human genetics and complex diseases.¹⁴ GWAS genotyping arrays can identify millions of SNPs across the genome in one assay; these variants are generally common and mostly not protein coding. Effect sizes of SNPs associated with disease are mostly small, requiring very large study samples to reach statistical significance.^13,14 Fortunately, technological advances have lowered the costs of genotyping considerably over the years,⁶⁷ and GWAS on hundreds of thousands of individuals are becoming more common.

8.1 GWAS of Refractive Errors and Myopia

GWAS for myopia have been performed using myopia as a dichotomous outcome or refractive error as a quantitative trait. Several endophenotypes have also been considered: spherical equivalent, axial length, corneal curvature, and age of diagnosis of myopia.

Figure 2 provides an overview of all associated loci and nearby genes, their frequency, and effect sizes.

Figure 2.

Effect sizes of common and rare variants for myopia and refractive error. Overview of SNPs and annotated genes found in the most recent GWAS meta-analysis.²⁴ The x-axis displays the minor allele frequency of each SNP; y-axis displays the effect size of the individual SNP in diopters; We transformed the z-scores of the fixed effect meta-analysis between CREAM (refractive error) and 23andMe (age of diagnosis of myopia) into effect sizes in diopters with the following formula²⁴: .

8.1.1 Myopia Case-Control Design

The case-control design using (high) myopia as a dichotomous outcome has been especially popular in East Asia. The first case-control GWAS was performed in a Japanese cohort in 2009.⁶⁸ It comprised 830 cases of pathologic myopia (defined as axial length >26 mm) and 1911 controls from the general population. The strongest association was located at 11q24.1, approximately 44 kb upstream of the BH3-like motif containing, cell death inducer (BLID) gene, and conferred odds of higher myopia of 1.37 (95% confidence interval [CI] 1.21–1.54). Subsequently, a GWAS meta-analysis of two ethnic Chinese cohorts was performed in 287 cases of high myopia (defined as ≤ −6 diopters [D]) and 673 controls. The strongest association was for an intronic SNP within the catenin delta 2 (CTNND2) gene on 5p15.2.⁶⁹ Neither of these associations met the conventional GWAS threshold (P ≤ 5 × 10⁻⁸) for statistical significance due to small sample size. Nevertheless, the locus at 5p15 encompassing the CTNND2 gene was later confirmed by other Asian studies.^70–72

Li et al.⁷³ studied 102 high myopia cases (defined as ≤ −8 D with retinopathy) and 335 controls in an ethnic Chinese population. The strongest association (P = 7.70 × 10⁻¹³) was a high-frequency variant located in a gene desert within the MYP11 myopia linkage locus on 4q25. In a similar ethnic Han Chinese population of 419 high myopia cases (≤ −6 D) and 669 controls, Shi et al.^73,74 identified the strongest association (P = 1.91 × 10⁻¹⁶) at an intronic, high-frequency variant within the mitochondrial intermediate peptidase (MIPEP) gene on 13q12. Neither hit has been replicated, even in studies with similar design, phenotypic definition, and ethnic background.

In 2013, two papers reported loci for high myopia in Asian populations and these were successfully replicated. Shi et al.⁷⁵ studied a Han Chinese population of 665 cases with high myopia (≤ −6 D) and 960 controls. Following two-stage replication in three independent cohorts, the most significantly associated variant (P = 8.95 × 10⁻¹⁴) was identified in the vasoactive intestinal peptide receptor 2 (VIPR2) gene within the MYP4 locus, followed by three other variants within a linkage disequilibrium block in the syntrophin beta 1 (SNTB1) gene (P = 1.13 × 10⁻⁸ to 2.13 × 10⁻¹¹). Khor et al.⁷⁶ reported a meta-analysis of four GWAS including 1603 cases of “severe” myopia and 3427 controls of East Asian ethnicity. After replication and meta-analysis, the SNTB1 gene was confirmed, and a novel variant within the ZFHX1B gene (also known as zinc finger E-box binding homeobox 2 [ZEB2]) reached genome-wide significance (P = 5.79 × 10⁻¹⁰).

In 2018, a pathologic myopia case-control study was performed in cohorts of Asian ancestry, using participants with −5.00 D or more myopia with an axial length >26 mm. Fundus photographs were graded pathologic or nonpathologic (N_cases = 828, N_controls = 3624). The researchers found a novel genetic variant in the coiled-coil domain containing 102B (CCDC102B) locus (P = 1.46 × 10⁻¹⁰), which was subsequently replicated in an independent cohort (P = 2.40 × 10⁻⁶). This gene is strongly expressed in the RPE and choroid. As myopic maculopathy is the primary cause of blindness in high myopia, further functional investigation could be valuable.⁷⁷

In Europe, a French case-control GWAS was performed on 192 high myopia cases (≤ −6 D) and 1064 controls, and a suggestive association was identified within the MYP10 linkage locus, 3 kb downstream of protein phosphatase 1 regulatory subunit 3B (PPP1R3B). However, this association did not reach genome-wide statistical significance, and no previously reported loci were replicated.⁷⁸ Later, in 2016, the direct-to-consumer genetic testing company 23andMe (Mountain View, CA, USA) published a large GWAS on self-reported myopia (N_cases = 106,086 and N_controls = 85,757; all European ancestry), and identified more than 100 novel loci for myopia.⁷⁹ Because this study was intended for association analyses between traits, precise locus definitions, post-GWAS quality control, and replication were not performed.

8.1.2 Quantitative Design on Spherical Equivalent

Studies that considered refractive error as a quantitative trait, and included subjects from the general population who displayed the entire range of refractive error, have been more successful. In 2010, the first GWAS for spherical equivalent were carried out in two European populations: a British cohort of 4270 individuals and a Dutch cohort of 5328 individuals.^80,81 Two loci surpassed the GWAS threshold and were replicated: one near the RASGFR1 gene on 15q25.1 (P = 2.70 × 10⁻⁰⁹) and the other near GJD2 on 15q14 (P = 2.21 × 10⁻¹⁴). Subsequently, a meta-analysis was performed on 7280 individuals with refractive error from five different cohorts, which included various ethnic populations across different continents, and findings were replicated in 26,953 samples. A novel locus including the RBFOX1 gene on chromosome 16 reached genome-wide significance (P = 3.9 × 10⁻⁹).⁸²

These collaborations paved the way for the formation of a large consortium to achieve higher statistical power for gene finding. CREAM was established in 2010 and included researchers and cohorts from the United States, Europe, Asia, and Australia. Its first collaborative work was replication of SNPs in the previously identified 15q14 loci.⁸³ Other studies followed this approach, and confirmed 15q14 as well as the 15q25 locus.^84,85 Subsequently, CREAM conducted a GWAS meta-analysis based on HapMapII imputation⁸⁶ with 35 participating studies comprising 37,382 individuals of European descent and 12,332 of Southeast Asian ancestry with data on GWAS and spherical equivalent. This study enabled replication of GJD2, RASGRF1, and RFBOX1 and identification of 23 novel loci at genome-wide significance: BICC1, BMP2, CACNA1D, CD55, CHD7, CHRNG, CYP26A1, GRIA4, KCNJ2, KCNQ5, LOC100506035, LAMA2, MYO1D, PCCA, TJP2, PTPRR, SHISA6, PRSS56, RDH5, RORB, SIX6, TOX, and ZMAT472.⁸⁷

Meanwhile, 23andMe performed a contemporaneous large GWAS on 55,177 individuals of European descent by using a survival analysis, based on the first release of 1000G⁸⁸ (a catalog of human genetic variation). Its analysis was based on self-reported presence of myopia and age of spectacle wear as a proxy for severity. 23andMe also replicated GJD2, RASGRF1, and RFBOX1 and identified 11 new loci: BMP3, BMP4, DLG2, DLX1, KCNMA1, LRRC4C, PABPCP2, PDE11A, RGR, ZBTB38, ZIC2.⁸⁹ Of the 22 loci discovered by CREAM, 8 were replicated by 23andMe, and 16 of the 20 loci identified by 23andMe were confirmed by CREAM. This was surprising, as the studies used very different phenotyping methods. In addition, the effect sizes of 25 loci were very similar, despite analyses on different scales: diopters for CREAM and hazard ratios for 23andMe.⁹⁰ After these two publications, replication studies provided validation for KCNQ5, GJD2, RASGRF1, BICC1, CD55, CYP26A1, LRRC4C, LAMA2, PRSS56, RFBOX1, TOX, ZIC2, ZMAT4, and B4GALNT2 in per-SNP analyses, and for GRIA4, BMP2, BMP4, SFRP1, SH3GL2, and EHBP1L1 in gene-based analyses.^91–96

Although CREAM and 23andMe found a large number of loci, only approximately 3% of the phenotypic variance of refractive error was explained.^87,89 Larger GWAS meta-analyses were clearly needed, and the two large studies combined efforts. This new GWAS meta-analysis was based on the phase 1 version 3 release of 1000G, included 160,420 participants, and findings were replicated in the UK Biobank (95,505 participants). Using this approach, the number of validated refractive error loci increased to 161. A high genetic correlation between European and Asian individuals (>0.78) was found, implying that the genetic architecture of refractive error is quite similar for European and Asian individuals. Taken together, these genetic variants accounted for 7.8% of the explained phenotypic variance, leaving room for improvement. Even so, polygenic risk scores, which are constructed by the sum of effect sizes of all risk variants per individual depending on their genotypes, were well able to distinguish individuals with hyperopia from those with myopia at the lower and higher deciles. Interestingly, those in the highest decile had a 40-fold greater risk of myopia. The predictive value (area under the curve) of these risk scores for myopia versus hyperopia, adjusted for age and sex, was 0.77 (95% CI 0.75–0.79).

The next step will include GWAS on even larger sample sizes. Although this will improve the explained phenotypic variance, it is unlikely that GWAS will uncover the entire missing heritability. SNP arrays do not include rare variants, nor do they address gene-environment and gene-gene interactions, or epigenetic effects.

8.1.3 GWAS on Refractive Error Endophenotypes

As myopia is mostly due to increased axial length, researchers have used this parameter as a myopia proxy or “endophenotype.” The first axial length GWAS examined 4944 individuals of East and Southeast Asian ancestry, and a locus on 1q41 containing the zinc finger pseudogene ZC3H11B reached genome-wide significance (P = 4.38 × 10⁻¹⁰).^82,97 A much larger GWAS meta-analysis of axial length comprised 12,531 European individuals and 8216 Asian individuals.⁹³ This study identified eight novel genome-wide significant loci (RSPO1, C3orf26, LAMA2, GJD2, ZNRF3, CD55, MIP, ALPPL2), and also replicated the ZC3H11B gene. Notably, five of these loci had been associated with refractive error in previous GWAS.

Several relatively small GWAS have been performed for corneal curvature, and identified associations with FRAP1, PDGFRA (also associated with eye size), CMPK1, and RBP3.^93,98–101 More recently Miyake et al.^101,102 published a two-stage GWAS for three myopia-related traits: axial length, corneal curvature, and refractive error. The study was performed in 9804 Japanese individuals, with trans-ethnic replication in Chinese and Caucasian individuals. A novel gene, WNT7B, was identified for axial length (P = 3.9 × 10⁻¹³) and corneal curvature (P = 2.9 × 10⁻⁴⁰), and the previously reported association with GJD2 and refractive error was replicated.

8.2 Genome-Wide Pathway Analyses

The main goal of GWAS is to improve insight on the molecules involved in disease, and help identify disease mechanisms. For myopia, a retina-to-sclera signaling cascade had been proposed for many years (see accompanying paper IMI – Report on Experimental Models of Emmetropization and Myopia¹⁰³), but knowledge on its molecular drivers was limited. Several attempts were made to translate the findings from refractive error GWAS into this cascade.^87,89,104 Here we provide an overview of genes annotated to the risk variants and their relationship to the underlying biological mechanism.

Deducted from the CREAM GWAS, pathways included neurotransmission (GRIA4), ion transport (KCNQ5), retinoic acid metabolism (RDH5), extracellular matrix remodeling (LAMA2, BMP2), and eye development (SIX6, PRSS56). Likewise, 23andMe proposed extracellular matrix remodeling (LAMA2, ANTXR2), the visual cycle (RDH5, RGR, KCNQ5), neuronal development (KCNMA1, RBFOX1, LRRC4C, NGL-1, DLG2, TJP2), eye and body growth (PRSS56, BMP4, ZBTB38, DLX1), and retinal ganglion cells (ZIC2, SFRP1)¹⁰⁵ as functions. Hysi et al.¹⁰⁶ performed pathway analyses using both the CREAM and 23andMe GWAS, and reported that plasma membrane, cell-cell adhesion, synaptic transmission, calcium ion binding, and cation channel activity were significantly overrepresented in refractive error in two British cohorts. Furthermore, by examining known protein-protein interactions, the investigators identified that many genes are related to cell-cycle and growth pathways, such as the MAPK and TGF-beta/SMAD pathways.

The latest update on pathway analysis in myopia stems from the meta-GWAS from CREAM and 23andMe.²⁴ TGF-beta signaling pathway was a key player; the association with the DRD1 gene provided genetic evidence for a dopamine pathway. Most genes were known to play a role in the eye,¹⁰⁷ and most significant gene sets were “abnormal photoreceptor inner segment morphology” (Mammalian Phenotype Ontology [MP] 0003730; P = 1.79 × 10⁻⁷), “thin retinal outer nuclear layer” (MP 0008515), “detection of light stimulus” (Gene Ontology [GO] 0009583), “nonmotile primary cilium” (GO 0031513), and “abnormal anterior-eye-segment morphology” (MP 0005193). Notably, RGR, RP1L1, RORB, and GNB3 were present in all of these meta-gene sets. Taken together, retinal cell physiology and light processing are clearly prominent mechanisms for refractive error development, and all cell types of the neurosensory retina, RPE, vascular endothelium, and extracellular matrix appear to be involved (Fig. 3). Novel mechanisms included rod-and-cone bipolar synaptic neurotransmission, anterior-segment morphology, and angiogenesis.²⁴

Figure 3.

Schematic overview of expression in retinal cells of refractive error and syndromic myopia genes according to literature. Bold: genes identified for both common refractive error and in syndromic myopia.

9. Whole-Exome and Whole-Genome Sequencing

Unlike GWAS, whole-exome sequencing (WES) and whole-genome sequencing (WGS) have the potential to investigate rare variants. Exomes are interesting, as they directly contribute to protein translation, but they constitute only approximately 1% of the entire genome. WGS allows for identification of variants across the entire genome, but requires a high-throughput computational infrastructure and remains costly.

WES has been conducted primarily in case-control studies of early-onset high myopia or in specific families with a particular phenotype (i.e., myopic anisometropia) or inheritance pattern (i.e., X-linked).^108–111 Several novel mutations in known myopia genes were identified this way: CCDC111,¹⁰⁹ NDUFAF7,¹¹⁰ P4HA2,¹⁰⁸ SCO2,¹¹² UNC5D,¹¹¹ BSG,¹¹³ ARR3,¹¹⁴ LOXL3,¹¹⁵ SLC39A5,¹¹⁶ LRPAP1,¹¹⁷ CTSH,¹¹⁷ ZNF644.^118,119 Although most genetic variants displayed an autosomal dominant hereditary pattern,^{108,112,118,119} X-linked heterozygous mutations were identified in ARR3, only in female family members.¹¹⁴ The functions of these novel genes include DNA transcription (CCDC111, ZNF644), mitochondrial function (NDUFAF7, SCO2), collagen synthesis (P4HA2), cell signaling (UNC5D, BSG), retina-specific signal transduction (ARR3), TGF-beta pathway (LOXL3, SLC39A5, LRPAP1), and degradation of proteins in lysosomes (CTSH). Jiang et al.¹¹⁹ investigated family members with high myopia and identified new mutations in LDL receptor related protein associated protein 1 (LRPAP1), cathepsin H (CTSH ), zinc finger protein 644 isoform 1 (ZNF644), solute carrier family 39 (metal ion transporter) member 5 (SLC39A5), and SCO2, cytochrome c oxidase assembly protein (SCO2).¹¹⁹

Many clinicians have noticed that retinal dystrophies and ocular developmental disorders often coincide with myopia.¹¹⁵ This triggered Sun et al.¹²⁰ to evaluate variants in a large number of retinal dystrophy genes in early-onset high myopia in 298 unrelated myopia probands and their families, and they thereby identified 29 potentially pathogenic mutations in COL2A1, COL11A1, PRPH2, FBN1, GNAT1, OPA1, PAX2, GUCY2D, TSPAN12, CACNA1F, and RPGR, and most had an autosomal dominant inheritance pattern. Kloss et al.¹²¹ performed WES in 14 families with high myopia, and identified 104 new genetic variants located in both known MYP loci (e.g., AGRN, EME1, and HOXA2) and in new loci (e.g., ATL3 and AKAP12).

To date, WGS has not been conducted for myopia or refractive error, most likely due to the reasons mentioned above. When costs for WGS decrease, these studies will undoubtedly be conceived.

10. Gene-Environment Interaction

It has become clear that environmental factors are driving the recent epidemic rise in the prevalence of myopia.^122–126 To date, the most influential and consistent environmental factor is education. Studies have estimated that individuals going onto higher education have double the myopia prevalence compared with those who leave school after only primary education.^127–129 Education has been a primary focus for gene-environment (GxE) interaction analyses in myopia. GxE studies have the potential to show modification of the effect of risk variants by environmental exposures, but can also reveal genetic associations that were hidden in unexposed individuals.

One of the first GxE studies for myopia investigated variants in matrix metalloproteinases genes (MMP1–MMP10). Two SNPs (rs1939008 and rs9928731) that were first found to be associated with refraction in Amish families were also associated in a lower but not in the higher education group of the Age-Related Eye Disease Study (AREDS) study. These results suggest that variants in these genes may play a role in refractive variation in individuals not exposed to myopic triggers.^59,130 In contrast, a study combining human GWAS data and animal models of myopia provided an experimental example of GxE interaction involving a rare variant in the APLP2-gene only in children exposed to large amounts of daily reading.¹³¹ In addition, an analysis performed in five Singapore cohorts found risk variants in DNAH9, GJD2, and ZMAT4 that were more strongly associated in individuals who achieved higher secondary or university education.¹³² Significant biological interaction between education and other risk variants was studied using a genetic risk score of all known risk variants at the time (n = 26) derived from the CREAM meta-GWAS.¹³³ European subjects with a high genetic load in combination with university-level education had a far greater risk of myopia than those with only one of these two factors. A study investigating GxE interactions in children and the major environmental risk-factors, nearwork, time outdoors, and 39 SNPs derived from the CREAM meta-GWAS revealed nominal evidence of interaction with nearwork (top variant in ZMAT4).^133,134

GEWIS, using all variants from the CREAM meta-GWAS, revealed three novel loci (AREG, GABRR1, and PDE10A) for GxE in Asian populations, whereas no interaction effects were observed in Europeans due to many reasons, such as the quantitative differences in the intensity of nearwork during childhood.⁴⁸ Up to now, there is no robust evidence that there are fundamental differences in the genetic background of myopia risk between European and Asian individuals.

11. Mendelian Randomization

Mendelian randomization (MR) is a method that allows one to test or estimate a causal effect from observational data in the presence of confounding factors. MR is a specific type of instrumental variable analysis that uses genetic variants with well-understood effects on exposures or modifiable biomarkers.^135,136 Importantly, the SNP must affect the disease status only indirectly via its effect on the exposure of interest.¹³⁷ MR is particularly valuable in situations in which randomized controlled trials are not feasible, where it is applied to help elucidate biological pathways.

Currently, three studies have been published on MR in refractive error and myopia. The first, published in 2016, explored the effect of education on myopia.¹³⁸ This study constructed polygenic risk scores of genetic variants found in GWAS for educational attainment, and used these as the instrumental variable. Subsequently, results of three cohorts (Cooperative Health Research in the Region Augsburg [KORA], AREDS, Blue Mountain Eye Study [BMES]; total N = 5649) were meta-analyzed. Strikingly, approximately 2 years of education was associated with a myopic shift of −0.92 ± 0.29 D (P = 1.04 × 10⁻³), which was even larger than the observed estimate. Similar results were observed in data from the UK Biobank study (N = 67,798); MR was performed and causality of education was tested for myopic refractive error bi-directionally.¹³⁹ Genetic variants for years of education from Social Science Genetic Association Consortium (SSGAC) and 23andMe studies were considered. Analyses of the observational data suggested that every additional year of education was associated with a myopic shift of −0.18 D per year (95% CI −0.19 to −0.17; P < 2.0⁻¹⁶). MR suggested the true causal effect was stronger: −0.27 D per year (95% CI −0.37 to −0.17; P = 4.0⁻⁸). As expected, there was no evidence that myopia was a cause for education (P = 0.6). The conclusion from these studies was that education appears truly causally related to myopia, and effects calculated by the current observational studies may be underestimated.

Because several studies had proposed that vitamin D has a protective effect against myopia,^140–142 the third MR study investigated the causality of low vitamin D concentrations on myopia. Genetic variants of the DHCR7, CYP2R1, GC, and CYP24A1 genes with known effects on serum levels of vitamin D were used as instrumental variables in a meta-analysis of refractive error in CREAM (N_EUR = 37,382 and N_ASN = 8,376). The estimated effects of vitamin D on refractive error were small in both ethnicities (Caucasians: −0.02 [95% CI −0.09, 0.04] D per 10 nmol/L increase in vitamin D concentration; Asian individuals: 0.01 [95% CI −0.17, 0.19] D per 10 nmol/L increase).These results suggest that the causal effect of vitamin D on myopia is very small, if any. Therefore, associations with vitamin D levels in the observational studies are likely to represent the effect of time spent outdoors.

12. Epigenetics

Epigenetic changes refer to functionally relevant changes to the genome that do not involve the nucleotide sequence of DNA. They represent other changes of the helix structure, such as DNA methylation and histone modification,¹⁴³ and these changes can regulate gene expression. Noncoding RNAs are small molecules that can also regulate gene expression, mainly at the posttranscriptional level; they can be epigenetically controlled but can also drive modulation of the DNA chromatin structure themselves.¹⁴⁴ Investigations into epigenetic changes of eye diseases still face some important technological hurdles. High-throughput next-generation sequencing technologies and high-resolution genome-wide epigenetic profiling platforms are still under development, and accessibility of RNA expression in human ocular tissues¹⁴⁵ is limited. Moreover, epigenetic changes are tissue- and time-specific, so it is essential to study the right tissue at the correct developmental stage. Animal models are often used as a first step before moving to humans, although epigenetic processes are not always conserved across species. Nevertheless, there have been some attempts to reveal epigenetic changes involved in myopia development.

A experiment using monocular form deprivation in a mouse model found that hypermethylation of CpG sites in the promoter/exon 1 of COL1A1 may underlie reduced collagen synthesis at the transcriptional level in myopic scleras.¹⁴⁶ A human study analyzing myopic individuals found that methylation of the CpG sites of the CRYAA promotor leads to lower expression of CRYAA in human lens epithelial cells.¹⁴⁷

Myopia studies evaluating the role of noncoding RNAs are more common. The latest GWAS meta-analysis found 31 loci residing in or near regions transcribing small noncoding RNAs, thus hinting toward the key role of posttranscriptional processes and epigenetic regulation.^24,144 MicroRNAs (miRNAs) are the best-characterized family of small noncoding RNAs. In their mature form, they are approximately 19 to 24 nucleotides in length and regulate hundreds of genes. They are able to bind to 3′ untranslated regions (UTRs) on RNA polymers by sequence-specific posttranscriptional gene silencing; one miRNA can regulate the translation of many genes. miRNAs have been a hot topic in past years due to the potential clinical application of these small RNA sequences: accessibility of the retina for miRNA-based therapeutic delivery has great potential for preventing and treating retinal pathology.¹⁴⁸ In a case-control study, Liang et al.¹⁴⁹ identified a genetic variant, rs662702, that was associated with the risk of extreme myopia in a Taiwanese population. The genetic variant was located at the 3′-UTR of PAX6, which is decreased in myopia. rs662702 is localized near the seed region of miR-328, and the C > T substitution leads to a mismatch between miR-328 and PAX6 mRNA. Further functional study indicated that the risk C allele reduced PAX6 expression relative to the T allele, which could result from knockdown effect of the C allele by miR-328. Therefore, reducing miR-328 may be a potential strategy for preventing or treating myopia.⁶¹ Another study focused on miR-184. This miRNA is the most abundant one in the cornea and the crystalline lens, and sequence mutations have been associated with severe keratoconus with early-onset anterior polar cataract. Lechner et al.^149,150 sequenced miR-184 in 96 unrelated Han southern Chinese patients with axial myopia, but no mutations were detected. Xie et al.¹⁵¹ analyzed rs157907 A/G in miR-29a and rs10877885 C/T in let-7i in a severe myopia case-control study (N_cases = 254; N_controls = 300). The G allele of the rs157907 locus was significantly associated with decreased risk of severe myopia (P = 0.04), launching the hypothesis that rs157907 A/G might regulate miR-29a expression levels. Functional studies are needed to provide evidence for this theory.

13. Concluding Remarks

Research on myopia genetics, genetic epidemiology, and epigenetics is flourishing and is providing a wealth of new insights into the molecules involved in myopiagenesis. Despite this progress, the chain of events forming the myopia-signaling cascade and the triggers for scleral remodeling are still largely unknown. Next steps should include all the novel technological advances for dissecting complex disorders, such as expansion of omics (such as genomics, transcriptomics, proteomics, and metabolomics), using multisource study populations, environmental genomics, and systems biology to organically integrate findings and improve our understanding of myopia development in a quantitative way via big data analytics (i.e., combining multi-omics and other approaches with deep learning or artificial intelligence). Expanding our knowledge of pathologic mechanisms and ability to pinpoint at-risk individuals will lead to new therapeutic options, better patient management, and, ultimately, prevention of complications and visual impairment from myopia.

Appendix

The CREAM Consortium

Joan E. Bailey-Wilson,¹ Paul Nigel Baird,² Amutha Barathi Veluchamy,^3–5 Ginevra Biino,⁶ Kathryn P. Burdon,⁷ Harry Campbell,⁸ Li Jia Chen,⁹ Ching-Yu Cheng,^10–12 Emily Y. Chew,¹³ Jamie E. Craig,¹⁴ Phillippa M. Cumberland,¹⁵ Margaret M. Deangelis,¹⁶ Cécile Delcourt,¹⁷ Xiaohu Ding,¹⁸ Cornelia M. van Duijn,¹⁹ David M. Evans,^20–22 Qiao Fan,²³ Maurizio Fossarello,²⁴ Paul J. Foster,²⁵ Puya Gharahkhani,²⁶ Adriana I. Iglesias,^19,27,28 Jeremy A. Guggenheim,²⁹ Xiaobo Guo1,^8,30 Annechien E. G. Haarman,^19,28 Toomas Haller,³¹ Christopher J. Hammond,³² Xikun Han,²⁶ Caroline Hayward,³³ Mingguang He,^2,18 Alex W. Hewitt,^2,7,34 Quan Hoang,^3,35 Pirro G. Hysi,³² Robert P. Igo Jr.,³⁶ Sudha K. Iyengar,^36–38 Jost B. Jonas,^39,40 Mika Kähönen,^41,42 Jaakko Kaprio,^43,44 Anthony P. Khawaja,^25,45 Caroline C. W. Klaver,^19,28,46 Barbara E. Klein,⁴⁷ Ronald Klein,⁴⁷ Jonathan H. Lass,^36,37 Kris Lee,⁴⁷ Terho Lehtimäki,^48,49 Deyana Lewis,¹ Qing Li,⁵⁰ Shi-Ming Li,⁴⁰ Leo-Pekka Lyytikäinen,^48,49 Stuart MacGregor,²⁶ David A. Mackey,^2,7,34 Nicholas G. Martin,⁵¹ Akira Meguro,⁵² Andres Metspalu,³¹ Candace Middlebrooks, Masahiro Miyake,⁵³ Nobuhisa Mizuki,⁵² Anthony Musolf,¹ Stefan Nickels,⁵⁴ Konrad Oexle,⁵⁵ Chi Pui Pang,⁹ Olavi Pärssinen,^56,57 Andrew D. Paterson,⁵⁸ Norbert Pfeiffer,⁵⁴ Ozren Polasek,^59,60 Jugnoo S. Rahi,^1,5,25,61 Olli Raitakari,^62,63 Igor Rudan,⁸ Srujana Sahebjada,² Seang-Mei Saw,^64,65 Dwight Stambolian,⁶⁶ Claire L. Simpson,^1,67 E-Shyong Tai,⁶⁵ Milly S. Tedja,^19,28 J. Willem L. Tideman,^19,28 Akitaka Tsujikawa,⁵³ Virginie J. M. Verhoeven,^19,27,28 Veronique Vitart,³³ Ningli Wang,⁴⁰ Juho Wedenoja,^43,68 Wen Bin Wei,⁶⁹ Cathy Williams,²² Katie M. Williams,³² James F. Wilson,^8,33 Robert Wojciechowski1,^70,71 Ya Xing Wang,⁴⁰ Kenji Yamashiro,⁷² Jason C. S. Yam,⁹ Maurice K. H. Yap,⁷³ Seyhan Yazar,³⁴ Shea Ping Yip,⁷⁴ Terri L. Young,⁴⁷ Xiangtian Zhou⁷⁵

¹Computational and Statistical Genomics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland, United States

²Centre for Eye Research Australia, Ophthalmology, Department of Surgery, University of Melbourne, Royal Victorian Eye and Ear Hospital, Melbourne, Australia

³Singapore Eye Research Institute, Singapore National Eye Centre, Singapore

⁴DUKE-NUS Medical School, Singapore, Singapore

⁵Department of Ophthalmology, National University Health Systems, National University of Singapore, Singapore

⁶Institute of Molecular Genetics, National Research Council of Italy, Pavia, Italy

⁷Department of Ophthalmology, Menzies Institute of Medical Research, University of Tasmania, Hobart, Australia

⁸Centre for Global Health Research, Usher Institute for Population Health Sciences and Informatics, University of Edinburgh, Edinburgh, United Kingdom

⁹Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Hong Kong Eye Hospital, Kowloon, Hong Kong

¹⁰Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore

¹¹Ocular Epidemiology Research Group, Singapore Eye Research Institute, Singapore National Eye Centre, Singapore

¹²Ophthalmology & Visual Sciences Academic Clinical Program (Eye ACP), DUKE-NUS Medical School, Singapore

¹³Division of Epidemiology and Clinical Applications, National Eye Institute/National Institutes of Health, Bethesda, Maryland, United States

¹⁴Department of Ophthalmology, Flinders University, Adelaide, Australia

¹⁵Great Ormond Street Institute of Child Health, University College London, London, United Kingdom

¹⁶Department of Ophthalmology and Visual Sciences, John Moran Eye Center, University of Utah, Salt Lake City, Utah, United States

¹⁷Université de Bordeaux, INSERM, Bordeaux Population Health Research Center, Team LEHA, UMR 1219, F-33000 Bordeaux, France

¹⁸State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China

¹⁹Department of Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands

²⁰Translational Research Institute, University of Queensland Diamantina Institute, Brisbane, Queensland, Australia

²¹MRC Integrative Epidemiology Unit, University of Bristol, Bristol, United Kingdom

²²Department of Population Health Sciences, Bristol Medical School, Bristol, United Kingdom

²³Centre for Quantitative Medicine, DUKE-National University of Singapore, Singapore

²⁴University Hospital ‘San Giovanni di Dio,' Cagliari, Italy

²⁵NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, London, United Kingdom

²⁶Statistical Genetics, QIMR Berghofer Medical Research Institute, Brisbane, Australia

²⁷Department of Clinical Genetics, Erasmus Medical Center, Rotterdam, the Netherlands

²⁸Department of Ophthalmology, Erasmus Medical Center, Rotterdam, the Netherlands

²⁹School of Optometry & Vision Sciences, Cardiff University, Cardiff, United Kingdom

³⁰Department of Statistical Science, School of Mathematics, Sun Yat-Sen University, Guangzhou, China

³¹Institute of Genomics, University of Tartu, Tartu, Estonia

³²Section of Academic Ophthalmology, School of Life Course Sciences, King's College London, London, United Kingdom

³³MRC Human Genetics Unit, MRC Institute of Genetics & Molecular Medicine, University of Edinburgh, Edinburgh, United Kingdom

³⁴Centre for Ophthalmology and Visual Science, Lions Eye Institute, University of Western Australia, Perth, Australia

³⁵Department of Ophthalmology, Columbia University, New York, United States

³⁶Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, United States

³⁷Department of Ophthalmology and Visual Sciences, Case Western Reserve University and University Hospitals Eye Institute, Cleveland, Ohio, United States

³⁸Department of Genetics, Case Western Reserve University, Cleveland, Ohio, United States

³⁹Department of Ophthalmology, Medical Faculty Mannheim of the Ruprecht-Karls-University of Heidelberg, Mannheim, Germany

⁴⁰Beijing Tongren Eye Center, Beijing Tongren Hospital, Beijing Institute of Ophthalmology, Beijing Key Laboratory of Ophthalmology and Visual Sciences, Capital Medical University, Beijing, China

⁴¹Department of Clinical Physiology, Tampere University Hospital and School of Medicine, University of Tampere, Tampere, Finland

⁴²Finnish Cardiovascular Research Center, Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland

⁴³Department of Public Health, University of Helsinki, Helsinki, Finland

⁴⁴Institute for Molecular Medicine Finland FIMM, HiLIFE Unit, University of Helsinki, Helsinki, Finland

⁴⁵Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom

⁴⁶Department of Ophthalmology, Radboud University Medical Center, Nijmegen, the Netherlands

⁴⁷Department of Ophthalmology and Visual Sciences, University of Wisconsin-Madison, Madison, Wisconsin, United States

⁴⁸Department of Clinical Chemistry, Finnish Cardiovascular Research Center-Tampere, Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland

⁴⁹Department of Clinical Chemistry, Fimlab Laboratories, University of Tampere, Tampere, Finland

⁵⁰National Human Genome Research Institute, National Institutes of Health, Baltimore, Maryland, United States

⁵¹Genetic Epidemiology, QIMR Berghofer Medical Research Institute, Brisbane, Australia

⁵²Department of Ophthalmology, Yokohama City University School of Medicine, Yokohama, Kanagawa, Japan

⁵³Department of Ophthalmology and Visual Sciences, Kyoto University Graduate School of Medicine, Kyoto, Japan

⁵⁴Department of Ophthalmology, University Medical Center of the Johannes Gutenberg University Mainz, Mainz, Germany

⁵⁵Institute of Neurogenomics, Helmholtz Zentrum München, German Research Centre for Environmental Health, Neuherberg, Germany

⁵⁶Department of Ophthalmology, Central Hospital of Central Finland, Jyväskylä, Finland

⁵⁷Gerontology Research Center, Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland

⁵⁸Program in Genetics and Genome Biology, Hospital for Sick Children and University of Toronto, Toronto, Ontario, Canada

⁵⁹Gen-info Ltd, Zagreb, Croatia

⁶⁰University of Split School of Medicine, Soltanska 2, Split, Croatia

⁶¹Ulverscroft Vision Research Group, University College London, London, United Kingdom

⁶²Research Centre of Applied and Preventive Cardiovascular Medicine, University of Turku, Turku, Finland

⁶³Department of Clinical Physiology and Nuclear Medicine, Turku University Hospital, Turku, Finland

⁶⁴Myopia Research Group, Singapore Eye Research Institute, Singapore National Eye Centre, Singapore

⁶⁵Saw Swee Hock School of Public Health, National University Health Systems, National University of Singapore, Singapore

⁶⁶Department of Ophthalmology, University of Pennsylvania, Philadelphia, Pennsylvania, United States

⁶⁷Department of Genetics, Genomics and Informatics, University of Tennessee Health Sciences Center, Memphis, Tennessee, United States

⁶⁸Department of Ophthalmology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland

⁶⁹Beijing Tongren Eye Center, Beijing Key Laboratory of Intraocular Tumor Diagnosis and Treatment, Beijing Ophthalmology & Visual Sciences Key Lab, Beijing Tongren Hospital, Capital Medical University, Beijing, China

⁷⁰Department of Epidemiology and Medicine, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States

⁷¹Wilmer Eye Institute, Johns Hopkins Medical Institutions, Baltimore, Maryland, United States

⁷²Department of Ophthalmology, Otsu Red Cross Hospital, Nagara, Japan

⁷³Centre for Myopia Research, School of Optometry, The Hong Kong Polytechnic University, Hong Kong

⁷⁴Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong

⁷⁵School of Ophthalmology and Optometry, Eye Hospital, Wenzhou Medical University, China