MacGillivray, Thomas

Bernabeu Llinares, Miguel

Dhillon, Baljean

Yii, Fabian S.L.

Medical Research Council (MRC)

UK Research & Innovation (UKRI)

University of Edinburgh

2025-11-10T11:32:50Z

2025-11-10T11:32:50Z

2025-11-10

The seemingly innocuous, easily remediable symptom of myopia—blurred distance vision—belies an insidious reality: myopia predisposes the eye to a range of sight-threatening diseases, including, but not limited to, pathologic myopia (PM) and retinal detachment. With a rapidly rising prevalence where half of humanity is predicted to be myopic by 2050, we are witnessing not just the emergence of a bespectacled generation, but also an increasing burden of irreversible visual impairment caused by myopia-related diseases. My PhD focuses on two important knowledge gaps. First, the epidemiology of PM has never been systematically investigated in the UK. To address this, I estimated the prevalence of PM using a random sample of mid-life adults with high myopia (spherical equivalent refraction ≤ -5D) from the population-based UK Biobank cohort. The results revealed a prevalence of 41.7%—with most cases being relatively mild (diffuse chorioretinal atrophy); however, the age-related and progressive nature of PM means an elevated risk of irreversible visual impairment for those affected later in life. Second, existing clinical predictors of the risks of myopia-related diseases—notably spherical equivalent refraction (SER) and axial length (AL)—do not provide personalised information about the anatomy of the posterior eye. This limitation partly explains why, despite both metrics being strong predictors at the population level, substantial variations in individual risks remain uncaptured. In myopia, an excessive enlargement of the globe puts significant stress on (or stretches) important structures in the posterior segment, which may explain the increased susceptibility to diseases affecting the posterior eye. To address the second gap, I argue for a ‘fundus-centric’ view of myopia: the characterisation of myopia at a more anatomically relevant fundus level using computational methods—as a supplement to conventional on-axis (along the visual axis) descriptors of myopia (SER and AL). The underlying premise is that fundus imaging offers a valuable first approximation of the degree of ocular stretching in an individual eye, thus providing personalised information about the risks of myopia-related diseases. I first investigated how fundus features varied across a broad spectrum of SER in healthy eyes from 23,000 adults in the UK Biobank. A wide range of fundus features were found to vary non-linearly as SER progressed from hyperopia to myopia: the optic disc became larger, less circular and orientated with its superior pole tilting towards the fovea; the disc-fovea distance increased; the vasculature became less tortuous with decreased branching complexity; and the vascular arcades curved more inwardly towards the fovea. In myopia, these changes appeared to be exponential—consistent with the known exponential increase in the risks of PM and retinal detachment with increasing myopia. In another study, I observed an increase in neuroretinal rim pallor with decreasing SER, with the temporal horizontal/inferior rim exhibiting the steepest rate of increase—consistent with the expectation that the papillomacular nerve fibre bundle, linked to the temporal rim, is particularly susceptible to damage from ocular stretching due to its straighter course. Importantly, I found evidence indicating that fundus imaging reflected differences in posterior eye shape, even when SER, age and sex were held constant. These findings support the premise that fundus imaging contains information—beyond what is available from on-axis descriptor of myopia alone—about ocular changes that may collectively be relevant to risk stratification. This culminated in the development of a deep learning-based metric—fundus refraction offset (FRO)—intended as a summary of how ‘anatomically myopic’ a fundus appears. A more negative FRO value indicated a more ‘myopic looking’ fundus than typical for an eye with the same SER. In the first (cross-sectional) validation study, I found evidence that FRO reflected differences in macular thickness and choroidal vascularity index, both derived from optical coherence tomography, even after controlling for AL or SER, as well as age, sex and ethnicity. In the second (longitudinal) validation study, a more negative baseline FRO was found to be associated with an increased risk of retinal detachment or breaks over 12 years, even among individuals with similar baseline SER (AL information not available), age, sex, macular thickness and history of cataract surgery. To conclude, myopia severity can be further quantified at the fundus level: such a ‘fundus-centric’ view of myopia allows the risks of complications to be captured at a level more personalised than is currently achievable using on-axis descriptors.

https://hdl.handle.net/1842/44151

http://dx.doi.org/10.7488/era/6675

en

The University of Edinburgh

Yii F, Nguyen L, Strang N, Bernabeu MO, Tatham AJ, MacGillivray T, Dhillon B. Factors Associated with Pathologic Myopia Onset and Progression: A Systematic Review and Meta-analysis. Ophthalmic Physiol Opt. 2024;44(5):963-76

Yii F, Bernabeu MO, Dhillon B, Strang N, MacGillivray T. Retinal Changes from Hyperopia to Myopia: Not all Diopters are Created Equal. Invest Ophthalmol Vis Sci. 2024;65(5):25.

Yii F, Gibbon S, MacGillivray T. Sectoral Changes in Neuroretinal Rim Pallor across Refractive Error. Ophthalmol Sci. 2025;5(3):100705.

Yii F, Strang N, Gibbon S, MacGillivray T. Can Fundus Features Tell us Something about 3D Eye Shape? Ophthalmic Physiol Opt. 2025.

Yii F, MacCormick IJ, Strang N, Bernabeu MO, MacGillivray T. Fundus Refraction Offset as an Individualized Myopia Biomarker. JAMA Ophthalmol. 2025

Yii F, Strang N, Moulson C, Dhillon B, Bernabeu MO, MacGillivray T. The optical nature of myopic changes in retinal vessel calibre. Ophthalmology Science. 2024:100631

Yii F, Strang N, Bernabeu MO, Dhillon B, MacGillivray T, MacCormick IJC. Epidemiology of pathologic myopia in UK adults with high myopia. medRxiv. 2024:2024.11.19.24317505. 5. Zhang H, Cisse M, Dauphin YN, Lopez-Paz D. mixup: Beyond Empirical Risk Minimization2017 October 01, 2017:[arXiv:1710.09412 p.]. Available from: https://ui.adsabs.harvard.edu/abs/2017arXiv171009412Z.

Yii FS. Emmetropic eye growth in East Asians and non-East Asians. Ophthalmic Physiol Opt. 2023;43(6):1412-8. Epub 20230627.

Yii F. Cultural roots of the myopia boom in Confucian Asia and their implications. J Public Health Policy. 2024;45(4):786-98. Epub 20240810.

Yii FS, He M, Chappell F, Bernabeu M, MacGillivray T, Dhillon B, et al. Higher intraocular pressure is associated with slower axial growth in children with non-pathological high myopia. Eye. 2024;38:1208-14.

Yii F, Strang N, Bernabeu MO, Dhillon B, MacGillivray T. Corneal biomechanics are not exclusively compromised in high myopia. Ophthalmic Physiol Opt. 2024;44(5):977-86. Epub 20240402

Yii F, Nguyen L, Strang N, Bernabeu MO, Tatham AJ, MacGillivray T, et al. Factors associated with pathologic myopia onset and progression: A systematic review and meta-analysis. Ophthalmic Physiol Opt. 2024;44(5):963-76.

Gibbon S, Muniz-Terrera G, Yii FSL, Hamid C, Cox S, Maccormick IJC, et al. PallorMetrics: Software for Automatically Quantifying Optic Disc Pallor in Fundus Photographs, and Associations With Peripapillary RNFL Thickness. Transl Vis Sci Technol. 2024;13(5):20

C​r​e​a​t​i​v​e ​C​o​m​m​o​n​s: ​A​t​t​r​i​b​u​t​i​o​n 4.0 International (​C​C-​B​Y 4.0)

https://creativecommons.org/licenses/by/4.0/

myopia

data science

retina

Fundus-centric view of myopia

A fundus-centric view of myopia

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy