The aim was to investigate the causal relationship between the lipidome and deep vein thrombosis (DVT) while identifying and quantifying the role of metabolites as potential mediators. Two-sample Mendelian randomization (MR) analysis of lipid species (n = 7174) and DVT (6767 cases and 330,392 controls) was performed using pooled data from genome-wide association studies. In addition, we quantified the proportion of metabolite-mediated lipidomic effects on DVT using 2-step MR. Phosphatidylcholine (18:0_18:2) levels mined from 179 lipids using MR analysis reduced the risk of DVT (odds ratio [OR]: 0.997
95% CI: 0.996-0.999
P = 4.25 × 10-4
false discovery rate [FDR] = 0.013). Octadecadienedioate (C18:2-DC) levels increased with the increasing phosphatidylcholine (18:0_18:2) levels (OR: 1.087, 95% confidence interval [CI]: [1.024, 1.154], P = .006). Octadecadienedioate (C18:2-DC) levels mined from 1400 metabolites using MR analysis reduced the DVT risk (OR: 0.997
95% CI: [0.996, 0.999], P = 6.11 × 10-6
FDR = 8.55 × 10-3). The proportion of the predicted genes for phosphatidylcholine (18:0_18:2) levels mediated by octadecadienedioate (C18:2-DC) levels was 7.8%. This study identified octadecadienedioate (C18:2-DC) levels as a potential mediator of the causal relationship between phosphatidylcholine (18:0_18:2) levels and DVT, which provides direction for the future investigation of DVT
however, further research on other potential mediators is still needed.