Type 2 diabetes (T2D) is a global epidemic causing major health and economic burden and significantly increases risk for cardiovascular disease (CVD). Due to the multifaceted nature of T2D, mechanisms underlying susceptibility to the disease are not fully elucidated. Compared to the genetics of T2D, environmental susceptibility to the disease is much less understood, explaining why accurate disease risk prediction is not feasible to reduce the growing prevalence of T2D. Recent genome-wide association studies (GWAS) discovered genetic associations with diet, a major environmental risk factor for T2D. Hence, to address such gap in the knowledge, we leveraged GWAS data for dietary preferences and examined their genetic associations with T2D including 21 related cardiometabolic traits. In Chapter 2, we first performed linkage disequilibrium score regression (LDSC) to estimate genetic correlations across dietary preferences, T2D and cardiometabolic traits and conducted colocalization experiments to identify genomic loci that underlie their genetic association. In Chapter 3 and 4, we utilized Mendelian randomization (MR) approach to evaluate causality in genetic relationships of diet with T2D and CVD. In Chapter 5, we employed directional clustering analysis to understand how the heterogeneity of obesity contributes to T2D risk and modulates the risk of cardiovascular outcomes. LDSC and colocalization analyses showed the shared genetic architecture across diet, T2D and cardiometabolic traits. MR detected the causal effects of specific dietary preferences on both T2D and CVD were substantially mediated through body mass index. In clustering analysis, metabolically healthy obesity subtype was not only associated with lower T2D risk but also with reduced risk for some CVDs, particularly artery disease. Overall, our findings provide genetic evidence for the association between diet and cardiometabolic disease (CMD). We discovered that diet interacts with body weight to causally influence CMD risk, underscoring that obesity is a strong etiological factor. Also, we observed the role of obesity heterogeneity in the variance of CMD risk. Together, our work demonstrates the importance of investigating largely environmental factors, which may partially have genetic component, in understanding the complexity in CMD risk.