Abstract: BACKGROUND: Age-related macular degeneration and deep vein thrombosis may share common pathophysiological mechanisms, but there is a lack of direct evidence regarding their relationship. Traditional studies are confounded by confounding factors and reverse causation.
OBJECTIVE: To investigate the causal relationship between age-related macular degeneration and deep vein thrombosis based on Mendelian randomization design. 
METHODS: Through a two-way Mendelian randomization analysis, single nucleotide polymorphisms of exposure and outcomes were obtained from publicly available genome-wide association studies, with deep vein thrombosis data from the FinnGen database in a European population with a sample size of 363 612 and 1 048 575 single nucleotide polymorphisms. In addition, we obtained data on age-related macular degeneration from the IEUOpenGWAS project, also from a European population sample of 105 248 cases covering 11 304 110 single nucleotide polymorphisms. In R4.4.1, we used the TwoSampleMR package (version 0.6.8) to explore the causal effects of exposure factors on outcomes. At the same time, we also conducted a sensitivity analysis via MR-Egger regression, weighted median, weighted model and simple model methods to ensure that the assessment results were robust and reliable. In addition, we used the “heterogeneity” function to test for heterogeneity, and the “horizontal pleiotropy” function and the MR-PRESSO test to further assess horizontal pleotropy. The Cochran’s Q test was used to determine whether there was statistical heterogeneity between single nucleotide polymorphisms, and the leave-one-out method was used to assess whether single nucleotide polymorphisms would significantly interfere with Mendelian randomization analysis. Funnel plots were drawn to assess the potential bias of single nucleotide polymorphisms. Forest plots were plotted to show the effect estimates of single nucleotide polymorphisms on exposure and outcomes, and their confidence intervals were plotted. Scatter plots were plotted to evaluate the relationship between the potency of single nucleotide polymorphisms and their causal effect size on outcome estimates. 
RESULTS AND CONCLUSION: Both forward and reverse studies showed that there was no causal association between age-related macular degeneration and the occurrence of deep vein thrombosis (P > 0.05). Sensitivity analysis showed that the main analysis results were reliable and robust, with no outliers, heterogeneity, and horizontal pleiotropy, and no single nucleotide polymorphism significantly affected the overall effect estimate. Although it is based on European population data, it has methodological reference value for Chinese biomedical research on complex disease associations. In this field, China can carry out multi-center large-sample studies, accurately analyze the internal links between Chinese population-related diseases, and provide a basis for prevention and treatment strategies and clinical practice.

Key words: age-related macular degeneration, deep vein thrombosis, Mendelian randomization, causal association, genome-wide association study data, FinnGen database, European population