Type 2 diabetes (T2D) predisposes to cardiovascular disease (CVD), but it is still unclear why some individuals with T2D are at disproportionately higher or lower risk. In this study, we employed a genetic stratification method to investigate the main clinical features that differ between two diabetogenic profiles associated concordantly with susceptibility for CVD or discordantly with protection against CVD. Quantifying concordant and discordant genetic predispositions improved CVD risk prediction, especially in men, correctly reassigning higher predicted risk to 5.4% of new male cases of MACE in UK Biobank. Moreover, higher genetically determined discordance reduced the risk associated with MACE in men. In-depth comparisons across a wide spectrum of phenotypes uncovered significant disparities between these two profiles. Subsequent causal inference analyses highlighted critical features of very-low-density lipoprotein particles influencing the discordance between T2D and CVD. We prioritized 8 distinct discordant genomic loci with potential protective effects traits against CVD in individuals with T2D. These findings provide clinically relevant valuable insights for personalized approaches to prevent and treat CVD in individuals with T2D.