Exploring the causal relationship between 16 eye diseases and stroke and their subtypes from a genome-wide perspective

Abstract

There is increasing evidence that eye diseases and stroke frequently co-occur, but the causal relationships remain elusive. Therefore, Mendelian randomization was used to investigate possible causal relationships between eye diseases and stroke (including its subtypes). This study utilized large-scale genome-wide association study pooled genetic data from two major databases: the IEU OpenGWAS project and the FinnGen databases. We then screened for instrumental variables that met the following three conditions: showing strong associations with the exposure factors, being independent of each other and independent of any confounders, and excluding instrumental variables to ensure that the F-value was greater than 10, and used the Inverse Variance Weighted (IVW) method to conduct causal analyses using the weighted median method, MR-Egger method, MR- PRESSO test and leave-one-out sensitivity test to test the robustness, heterogeneity and horizontal pleiotropy of the results. In order to control for the false positive rate in multiple hypothesis testing, a False Discovery Rate (FDR) correction was also performed. A total of 16 eye diseases and strokes and their subtypes were investigated in this study. In the IVW model, the MR study showed a total of 20 IVWs with p-values less than 0.05. We excluded 6 results with heterogeneity or pleiotropy by sensitivity analysis, and finally the following reliable results were left: (1) patients with age-related macular degeneration had a 5%, 2%, and 7% lower risk of subarachnoid hemorrhage, ischemic stroke, and small-vessel stroke; (2) patients with keratitis had a 12% higher risk of cardioembolic; (3) patients with optic atrophy had a 3% higher risk of stroke; (4) patients with amblyopia had a 3% higher risk of stroke; and (5) patients with other inflammation of eyelid had a small-vessel had a 20% elevated risk; (6) patients with ptosis of the eye had a 17% elevated risk of cardioembolic; (7) patients with strabismus have a 23% elevated risk of small-vessel; (8) patients with stroke had a refractive error by 17%; (9) patients with intracerebral hemorrhage had a 15% increased risk of uveitis; (10) patients with IS had an 11% increased risk of diabetic retinopathy; (11) patients with large-artery atherosclerosis had a 2% increased risk of glaucoma; (12) patients with cardioembolic had a 24% increased risk of amblyopia. From a genetic standpoint, keratitis, amblyopia, other eyelid inflammations, ptosis and strabismus are associated with increased risks of the risk of stroke or its subtypes. Conversely, age-related macular degeneration is associated with reduced risks of the risk of stroke or its subtypes. Furthermore, stroke or its subtypes increase the risks of refractive error, uveitis, diabetic retinopathy, glaucoma, and amblyopia. Nevertheless, it is imperative that these causal relationships be subjected to further verification through fundamental research and Randomized Controlled Trial (RCT).

Keywords: Causality; Eye diseases; Mendelian randomization; Stroke; Stroke subtypes.

Fig. 1

Study design and workflow. SNPs, single nucleotide polymorphisms. MR, Mendelian randomization. Sensitivity annalyses, including heterogeneity test, MR-Egger regression heterogeneity test, Cochrane’s Q test, and weighted median.

Fig. 2

Three assumptions about Instrumental variables (By Figdraw 2.0, https://www.figdraw.com/static/index.html#/). SNPs, single nucleotide polymorphisms.

Fig. 3

Summary of p-values for the inverse-variance weighted (IVW) MR method of predicting the relationship between eye diseases with stroke (including its subtypes). MR analysis of eye diseases as exposure and stroke (including its subtypes) as outcome. The p-value < 0.05 was deemed significant.

Fig. 4

Summary of p-values for the inverse-variance weighted (IVW) MR method of predicting the relationship between eye diseases with stroke (including its subtypes). MR analysis of stroke (including its subtypes) as exposure and eye diseases as outcome. The p-value < 0.05 was deemed significant.

Fig. 5

Main Mendelian randomization results forest plots. MR analysis of eye diseases as exposure and stroke (including its subtypes) as outcome.

Fig. 6

Main Mendelian randomization results forest plots. MR analysis of eye diseases as exposure and stroke (including its subtypes) as outcome. MR analysis of stroke (including its subtypes) as exposure and eye diseases as outcome.