Abstract
PURPOSE: To investigate the potential causal association between COVID-19 exposure and optic nerve and visual pathway disorders through a two-sample bidirectional Mendelian randomization (MR) analysis, and to provide empirical support for the lung-brain axis. METHODS: This MR analysis utilized publicly accessible summary-level data from genome-wide association studies on COVID-19 (n=158,783) and optic nerve and visual pathway diseases (n=412,181), primarily involving individuals of European descent. The random-effect inverse-variance weighted estimation was applied as the main analytical approach, complemented by MR-Egger, weighted median, and weighted mode methods. The heterogeneity and pleiotropy of the instrumental variables were assessed using Cochran's Q test, leave-one-out sensitivity analysis, MR-Egger intercept test, MR-PRESSO, and funnel plot evaluations. RESULTS: In the forward analysis, the inverse-variance weighted method identified a significant causal effect of COVID-19 on optic nerve and visual pathway disorders (odds ratio = 1.697, 95% confidence interval: 1.086-2.652, p = 0.020). Directionally consistent results were also observed with MR-Egger regression, weighted median, and weighted mode approaches. Conversely, the reverse analysis revealed no causal effects of optic nerve and visual pathway disorders on COVID-19 susceptibility. CONCLUSION: Our findings suggest that COVID-19 exposure may increase the risk of developing optic nerve and visual pathway disorders, supporting the lung-brain axis hypothesis. These results underscore the importance of vigilant monitoring of the visual system in patients recovering from COVID-19 and suggest potential avenues for future therapeutic strategies.