Abstract
Mendelian mutations in the Prothrombin gene (F2) and the factor V Leiden gene (F5) genes are established risk factors for venous thromboembolism (VTE). Walking pace is associated with the risk of coronary artery diseases, but no study has investigated its association with VTE. This study aimed to investigate the association and causality between walking pace and VTE, compare its population risk with established Mendelian mutations, and determine if blood biomarkers mediate its effect.We followed up 445,261 UK Biobank participants free of VTE at baseline. Self-reported walking pace was collected via touchscreen questionnaire at baseline. The carrier status of two Mendelian mutations in F2 and F5 genes was determined by the genotypes of rs1799963 (G20210A, c.*97 G > A) and rs6025 (p.R534Q), respectively. Cox proportional hazard model was used to estimate the effect of walking pace on incident VTE. We conducted a bidirectional Mendelian randomization (MR) analysis, by using 70 single-nucleotide polymorphisms (SNPs) from a walking pace genome-wide association studies (GWAS) and 93 SNPs from a VTE GWAS as instrumental variables. We used both individual-level data and GWAS summary statistics for mediation analysis.Over a median follow-up period of 12.8 years, 11,155 incident VTE cases were identified. The 10-year incidence rates for brisk and slow walking pace were 1.32% (confidence interval [CI]: 1.27-1.37%) and 3.90% (CI: 3.71-4.09%), respectively. For noncarriers, F2 and F5 carriers, the 10-year incidence rates were 1.70% (CI: 1.66-1.73%), 2.94% (CI: 2.66-3.22%), and 3.62% (CI: 3.39-3.84%), respectively. The overall risk of VTE for F5 mutation carriers with a brisk walking pace (2.65%) was smaller than that for noncarriers with a slow walking pace (3.66%). For F5 mutation carriers, brisk pace (but not steady pace) reduces the risk of VTE (p interaction < 0.05). MR analyses displayed a causal relationship (inverse variance weighted: p = 3.21 × 10(-5)) from walking pace to VTE incidence. Mediation analysis showed that serum albumin (ALB) and cystatin C (CYS) levels partially mediated the effect of brisk walking pace on the risk of VTE incidence, with mediation proportions of 8.7 to 11.7%, respectively.On the population scale, the protective effect of brisk walking pace offsets the risk of VTE caused by Mendelian mutations. We provided preliminary evidence that a brisk walking pace causally reduces the risk of VTE. Serum ALB and CYS partially mediate this effect.