Abstract
BACKGROUND AND AIMS: Susceptibility to cardiovascular disease (CVD) is driven by genetic and environmental risk factors. Diet is a modifiable and largely environmental risk factor for CVD. Genetic factors associated with a variety of dietary preferences revealed via recent genome-wide association studies (GWAS) allow further investigate the role of diet in liability to disease that has been limited to observational and epidemiologic studies with mixed findings. METHOD: We obtained publicly available genome-wide association data for 38 dietary preference traits and seven common CVDs to investigate causal hypotheses between diet as the exposure to CVD as outcomes using the statistical framework of Mendelian randomization (MR) for hypothesis testing and sensitivity analyses. We also conducted mediation analyses to evaluate the effects of dietary preferences on CVDs to elucidate potential causal graphs and estimate the effects of dietary preferences mediated by potential mediators. RESULTS: Across all methods, we identified 10 significant causal effects, which included eight dietary preferences across three CVD endpoints (Bonferroni-corrected P < 1.88 × 10(-4)). In sensitivity MR and mediation analysis, we observed that obesity - quantified by body mass index (BMI) - was a common mediator that contributed to many of these observed effects. We also found that educational attainment was an exclusive, additional mediator for the effect of preference for muesli with risk to peripheral artery disease (PAD). CONCLUSIONS: Our results provide genetic evidence for a link between diet and CVD that aligns with obesity-mediated risk of CVD in individuals in relation to their specific preferences for food.