Genetically predicted obesity and risk of deep vein thrombosis

Abstract

Aims/hypothesis: An association between obesity and deep vein thrombosis (DVT) has been revealed by observational studies, but it is not clear if the observed associations are causal, caused by confounding bias or reverse causation.

Methods: We performed a two-sample Mendelian Randomization (MR) study by obtaining exposure and outcome data from separate published studies. We utilized data from Genetic Investigation of Anthropometric Traits (GIANT, 339,224 participants) consortium and FinnGen project (FinnGen, 1785 DVT case and 84,462 control participants) to determine the causal effect of BMI on DVT.

Results: All three MR methods provided a positive association between BMI and DVT. Using IVW, we found evidence of causal relationships between BMI and DVT. BMI is positively associated with DVT (IVW odds ratio [OR] per SD increase in BMI = 1.67 [95% CI, 1.16-2.40]; P = 0.006). MR Egger and weighted median regression also showed directionally similar estimates (MR-Egger OR per SD increase in BMI, 2.50 [95% CI, 1.07-5.84], P = 0.034; weighted median OR per SD increase in BMI, 2.02 [95% CI, 1.10-3.71], P = 0.023). Both funnel plots and MR-Egger intercepts suggest no directional pleiotropic effects observed between BMI and DVT.

Conclusions/interpretation: Our findings provide evidence of significant causal association between BMI and DVT in agreement with observational studies. Taking measures to reduce the proportion of obesity may help reduce the incidence of DVT.

Keywords: Body mass index; Deep vein thrombosis; Mediation; Mendelian randomization.