Risk prediction to inform strategies for the primary prevention of cardiovascular disease is universally recommended by clinical practice guidelines in the US and worldwide. A risk-based prevention paradigm matches the intensity of the prevention effort with the absolute risk of the individual. Because the absolute risk reduction is directly proportional to the absolute risk of the individual (ie, individuals at higher predicted risk will experience greater benefit than those with a lower predicted risk from a given therapy), aligning the intensity of evidence-based prevention with absolute risk maximizes the benefit and efficiency of therapeutic interventions. Because of this paradigm for prevention, improving the accuracy and precision of risk assessment with novel markers or newer analytic methods is an active area of much investigation. The heritability of coronary heart disease (CHD) is estimated to be between 40% and 60%, so particular interest has been paid to the development of risk models that integrate inherited or genetic risk for CHD via polygenic risk scores (PRSs) in pursuit of the promise of precision medicine. The enthusiasm for such scores has been further propelled by the marked reductions in cost of genomic sequencing, advancements in statistical methods to analyze high-dimensional genomic data, and greater access to large diverse datasets with genomic information. Despite the absence of guideline recommendations for the use of PRSs in the primary prevention of cardiovascular disease, clinical, direct-to-consumer, and research use of PRSs for CHD has become widespread. However, the rapid pace with which these numerous PRSs have become available has left a critical gap in the understanding of the precision of individual PRSs and how to prioritize which PRS, if any, to use in risk prediction of CHD.
