Abstract
Observational studies have shown that serum iron status is related to heart failure (HF). However, due to confounding factors and reverse causation, there is still controversy over whether this association is causal. Our aim is to assess the causal relationship between genetically predicted serum iron status and HF. This study uses inverse variance weighting as the main method for Mendelian Randomization (MR) analysis, while also employing MR-Egger and weighted median as supplementary methods to evaluate the causal relationship between exposure factors and outcome variables. Additionally, the MR-Egger intercept test and MR-pleiotropy residual sum and outliers (MR-PRESSO) test were employed to assess the horizontal pleiotropy and outlier single nucleotide polymorphism. Cochran Q test in MR-Egger and inverse variance weighting methods was performed to evaluate the heterogeneity among the single nucleotide polymorphisms, and sensitivity analysis was performed by using the"leave-one-out"method. Finally, reverse MR analysis was used to verify the robustness of the results. Elevated serum ferritin levels (OR = 1.147, 95% CI 1.014-1.298, P-value = .029) have a causal relationship with a reduced risk of HF. However, there is no potential genetic causal relationship between serum iron, total iron-binding capacity, and transferrin saturation and HF (P >.05). In the reverse MR analysis, none of the 3 methods support reverse causality. The intercept test of the MR-Egger regression model, the MR-PRESSO test, and Cochran Q test all have P-values >.05, indicating that the selected single nucleotide polymorphisms do not exhibit horizontal pleiotropy or heterogeneity. In addition, the sensitivity analysis based on the leave-one-out method shows that a single single nucleotide polymorphism does not affect the robustness of the causal association effect value. There is a causal relationship between genetically predicted serum ferritin levels and the risk of HF, indicating that high levels of serum ferritin may be a protective factor against HF.