Causal relationship between matrix metalloproteinases and diabetic retinopathy: A bidirectional two-sample Mendelian...

Matrix metalloproteinases (MMPs) have been increasingly recognized as potential contributors or mediators in the pathogenesis of diabetic retinopathy (DR). Nevertheless, the relationships among MMPs and the risk of DR have not been definitively elucidated and still require further investigation. The bidirectional two-sample Mendelian randomization (MR) analyses were implemented to evaluate the causal effect of MMPs (specifically, MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, MMP-10, MMP-12, MMP-13, MMP-14, MMP-16, and MMP-17) on DR. The inverse variance weighted (IVW) served as the primary MR method, accompanied by sensitivity analyses to verify the result stability. In the forward MR analyses, genetically predicted MMP-7 was associated with an increased risk of DR (IVW: odds ratio = 1.15; 95% confidence interval: 1.05-1.25; P = 2.759 × 10⁻³]. In contrast, higher MMP-8 were robustly associated with a reduced risk of DR (IVW: odds ratio = 0.86; 95% confidence interval: 0.82-0.91; P = 5.506 × 10⁻⁸). No significant causal associations were observed between DR and the remaining MMPs. Sensitivity analyses revealed that the causal estimate for MMP-7 was sensitive to the threshold used for instrumental variables, whereas the association for MMP-8 was largely driven by 1 single nucleotide polymorphism (rs12614). In the reverse MR analyses, no significant causal effects of DR on any of the MMPs were detected after correction for multiple testing. This study provides preliminary and exploratory evidence supporting a causal association between MMPs and DR risk in European population. Specifically, MMP-7 may be associated with an increased risk of DR, while MMP-8 may be protective against DR. These findings suggest distinct pathophysiological roles for specific MMPs in DR development. However, given the exploratory nature of the analysis, further validation in mechanistic and longitudinal studies is warranted. not applicable.