BACKGROUND: Advances in high-throughput genomics facilitate the identification of novel genetic susceptibility variants for coronary heart disease (CHD). This may improve CHD risk prediction. The aim of the present simulation study was to investigate to what degree CHD risk can be predicted by testing multiple genetic variants (genetic profiling). METHODS: We simulated genetic profiles for a population of 100,000 individuals with a 10-year CHD incidence of 10%. For each combination of model parameters (number of variants, genotype frequency and odds ratio [OR]), we calculated the area under the receiver operating characteristic curve (AUC) to indicate the discrimination between individuals who will and will not develop CHD. RESULTS: The AUC of genetic profiles could rise to 0.90 when 100 hypothetical variants with ORs of 1.5 and genotype frequencies of 50% were simulated. The AUC of a genetic profile consisting of 10 established variants, with ORs ranging from 1.13 to 1.42, was 0.59. When 2, 5, and 10 times as many identical variants would be identified, the AUCs were 0.63, 0.69, and 0.76. CONCLUSION: To obtain AUCs similar to those of conventional CHD risk predictors, a considerable number of additional common genetic variants need to be identified with preferably strong effects.,
American Heart Journal
Erasmus MC: University Medical Center Rotterdam

van der Net, J., Janssens, C., Sijbrands, E., & Steyerberg, E. (2009). Value of genetic profiling for the prediction of coronary heart disease. American Heart Journal, 158(1), 105–110. doi:10.1016/j.ahj.2009.04.022