Background: Coronary heart disease, including its clinical manifestation, myocardial infarction (MI), is a common, complex disease with a substantive genetic component. State-of-the-art genetic epidemiology evaluates thousands of single nucleotide polymorphisms (SNPs) in association with disease cases and controls. In an independent but demographically similar population, this study tested 6 SNPs that were previously reported to be associated with MI.
Methods: Patients hospitalized for an acute MI (n = 413) at an early age (men < 55 years, women < 65 years) were compared with age-discordant (men > or = 65 years, women > or = 70 years) control patients (n = 792) who had no MI history and no hospitalization for MI at index angiography or during longitudinal follow-up. Six SNPs were genotyped in the genes palladin, ROS1, TAS2R50, OR13G1, and ZNF627.
Results: Findings were not different from the null hypothesis, with ZNF627 (AG vs. GG: odds ratio [OR] 1.47, P = .16; AA vs. GG: OR 1.20, P = .50) and both ROS1 SNPs (GG vs AA: OR 0.72, P = .21; CC vs GG: OR 0.74, P = .24) showing potentially interesting ORs but nonsignificant probabilities. After full adjustment for all SNPs and covariables, only the ZNF627 heterozygote genotype had OR > 1.5 (P = .14). Comparison of MI cases with controls without obstructive coronary artery disease and analyses stratified by sex provided similar findings.
Conclusions: Six SNPs previously reported to be associated with MI were not validated, suggesting that further investigation is needed to verify the applicability of those SNPs to cardiovascular medicine. These findings emphasize the high potential for false-positive results even in staged genome-wide association studies and further emphasize the need for continued refinement of cardiovascular genetic methodologies for clinical application.