Thyroid function and preeclampsia: a two-sample bidirectional Mendelian randomization study

Abstract

Background: Thyroid dysfunction has been associated with preeclampsia (PE) during pregnancy, but the observational results are conflicting. Our study aims to investigate the causal association and direction between genetically predicted effects of thyroid function on PE and vice versa via two large summary genetic data. Methods: We conducted a two-sample bidirectional Mendelian randomization (MR) study using genome-wide association studies (GWAS) summary data from two primarily European cohorts: the ThyroidOmics Consortium and the FinnGen Biobank. We applied the random effects inverse variance weighted (IVW) as our main analysis. MR-Egger and weighted median were used for sensitivity analysis. Statistical analysis was performed using the R program (version 4.3.0) with the two-sample package (version 0.5.6). Results: The results suggest that genetically predicted hyperthyroidism is causally associated with PE during pregnancy [β = 0.06, 95% confidence interval (CI): 1.01–1.12; P = 0.02], and genetically predicted hypothyroidism is also causally associated with PE during pregnancy (β = 0.11, 95% CI: 1.03–1.21; P = 0.01). These effects were further confirmed with sensitivity analysis. Conversely, preeclampsia is not associated with the risk of thyroid dysfunction in the reverse MR results: thyroid-stimulating hormone (β = 0.00, P = 0.92), free thyroxine (FT4) (β = −0.01, P = 0.56), triiodothyronine (FT3) (β = −0.00, P = 0.72), FT3/FT4 (β = −0.01, P = 0.38), thyroid peroxidase antibodies (β = −0.01, P = 0.64), hyperthyroidism (β = −0.11, P = 0.29) and hypothyroidism (β = 0.04, P = 0.12). Conclusion: Our study suggests that hyper-/hypo-thyroidism causally affected preeclampsia, while PE is not causally associated with thyroid dysfunctions.