Corneal Biomechanics as a Causal Factor in Myopia and Astigmatism: Evidence from Mendelian Randomization

Abstract

Purpose: The causal relationship between refractive errors and corneal biomechanical properties remains uncertain. This study aimed to clarify this relationship using Mendelian randomization (MR), offering new insights into the prevention and treatment of refractive errors.

Design: A bidirectional, 2-sample MR analysis.

Participants: Corneal biomechanical data were obtained from 97 653 European participants in the UK Biobank, whereas refractive error data were sourced from the UK Biobank and FinnGen consortia.

Methods: The exposures in this study were identified as corneal biomechanical parameters, specifically corneal hysteresis (CH) and the corneal resistance factor (CRF). The outcomes were defined as refractive errors, including myopia, hyperopia, and astigmatism, along with refractive power, encompassing both spherical and cylindrical power. A meta-analysis was performed to combine the MR estimates from both UK Biobank and FinnGen consortia, with heterogeneity assessed using the Q test and I² statistics. Additionally, a reverse MR analysis was conducted to examine the potential causal effect of the refractive status on corneal biomechanics.

Main outcome measures: Corneal hysteresis and CRF as causal factors in myopia and astigmatism.

Results: Data from UK Biobank revealed that CH and CRF were protective against the development of myopia (CH: odds ratio (OR) = 0.9936, P = 7.79 × 10^-4; CRF: OR = 0.9946, P = 2.41 × 10^-3) and astigmatism (CH: OR = 0.9975, P = 0.02; CRF: OR = 0.9977, P = 0.017). Conversely, increased corneal-compensated intraocular pressure was a risk factor for myopia development (OR = 1.0091, P = 2.07 × 10^{- 2}). The meta-analysis, which combined data from both sources, supported a causal relationship between CH and CRF and the development of myopia, although no significant causal link was found for hyperopia. Reverse MR analysis demonstrated a causal effect of spherical power on CH (OR = 1.0664, P = 4.32 × 10^{- 5}).

Conclusions: Corneal biomechanical parameters, particularly CH and CRF, may serve as early biomarkers for predicting myopia. This protective role supports their use in clinical screening to enhance early intervention strategies. Corneal-compensated intraocular pressure is a risk factor for myopia and represents a novel therapeutic target. Future research should clarify the underlying mechanisms and assess biomechanical interventions, potentially transforming refractive error management and reducing visual impairment.

Financial disclosures: The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Keywords: Astigmatism; Corneal biomechanics; Hyperopia; Mendelian randomization; Myopia.

Figure 1

Flowchart of the study design. CH = corneal hysteresis; CRF = corneal resistance factor; IOPcc = corneal-compensated intraocular pressure; IOPg = Goldmann-correlated intraocular pressure; SNPs = single nucleotide polymorphisms.

Figure 2

The figure shows the results of evaluating the causal effect of corneal biomechanical parameters on refractive errors (myopia and astigmatism) using 5 different methods of MR analysis (MR-Egger, weighted median, inverse variance weighting, simple mode, and weighted mode). The forest plot in the figure shows the ORs and their 95% CIs calculated by each method, with statistical significance determined by P values. Odds ratio values >1 are shown as risk factors, whereas OR values <1 are shown as protective factors. CI = confidence interval; IOPcc = corneal-compensated intraocular pressure; MR = Mendelian randomization; OR = odds ratio.

Figure 3

Scatterplots comparing the effects of SNPs on different parameters using 5 MR methods (inverse variance weighting, MR-Egger, simple mode, weighted median, and weighted mode). Each subplot reflects the relationship between SNP effects on the horizontal axis (corneal biomechanical parameters) and refractive status on the vertical axis. Black dots represent specific SNP effect estimates, and gray bands indicate the uncertainty of these estimates. The results of linear regression for the 5 methods are shown by lines of different colors. A, Single nucleotide polymorphism effects on CH and their association with myopia. B, Single nucleotide polymorphism effects on CRF and their association with myopia. C, Single nucleotide polymorphism effects on IOPcc and their association with myopia. D, Single nucleotide polymorphism effects on CH and their association with astigmatism. E, Single nucleotide polymorphism effects on CRF and their association with astigmatism. F, Single nucleotide polymorphism effects on CH and their association with spherical power. G, Single nucleotide polymorphism effects on CRF and their association with spherical power. H, Single nucleotide polymorphism effects on IOPcc and their association with spherical power. I, Single nucleotide polymorphism effects on CH and their association with cylindrical power. CH = corneal hysteresis; CRF = corneal resistance factor; IOPcc = corneal-compensated intraocular pressure; MR = Mendelian randomization; SNPs = single nucleotide polymorphisms.

Figure 6

The figure shows the results of evaluating the causal effect of corneal biomechanical parameters on refractive power (spherical and cylindrical power) using 5 different methods of MR analysis. The forest plot in the figure shows the beta coefficients (β) and their 95% CIs calculated by each method, with statistical significance determined by P values. β value >0 are shown as risk factors, whereas β value <0 are shown as protective factors. CI = confidence interval; MR = Mendelian randomization.

Figure 7

Causal analysis of CH and CRF on refractive status based on reverse MR analyses. A, Circular heatmap of MR IVW results among different traits. The P values of CH and CRF are shown from the outside to the inside. An asterisk indicates that the MR result is causal (P < 0.05). B, Scatter plots showing the SNP effects on spherical power and their association with CH. CH = corneal hysteresis; CRF = corneal resistance factor; IVW = inverse variance weighted; MR = Mendelian randomization; SNPs = single nucleotide polymorphisms.

Figure S

1Leave-one-out plot illustrating the causal effect of corneal biomechanics on refractive error. The red lines represent estimations derived from the inverse variance weighted (IVW) test. A, MR leave-one-out sensitivity analysis for corneal hysteresis (CH) on myopia. B, MR leave-one-out sensitivity analysis for corneal resistance factor (CRF) on myopia. C, MR leave-one-out sensitivity analysis for corneal-compensated intraocular pressure (IOPcc) on myopia. D, MR leave-one-out sensitivity analysis for CH on astigmatism. E, MR leave-one-out sensitivity analysis for CRF on astigmatism. F, MR leave-one-out sensitivity analysis for CH on spherical power. H, MR leave-one-out sensitivity analysis for CRF on spherical power. I, MR leave-one-out sensitivity analysis for IOPcc on spherical power. J, MR leave-one-out sensitivity analysis for CH on cylindrical power.

Figure S

2Funnel plot of the causal association between corneal biomechanics and refractive error. CHI = corneal hysteresis, CRF = corneal resistance factor, IOPcc = corneal-compensated intraocular pressure.