Objectives: Coronary artery disease (CAD) is often polygenic due to multiple mutations that contribute small effects to susceptibility. Since most prior studies only evaluated the contribution of single candidate genes, we therefore looked at a combination of genes in predicting early-onset CAD [apolipoprotein E (APOE) epsilon4, butyrylcholinesterase (BChE) K, peroxisome proliferator-activated receptor gamma2 (PPARgamma2) Pro12Ala and endothelial nitric oxide synthase (ENOS) T-786C].
Design and methods: We examined the frequencies, individually and in combination, of all four alleles among patients with early-onset CAD (n = 150; <50 years), late-onset CAD (n = 150; >65 years) and healthy controls (n = 150, age range 47-93 years). Differences in the proportion of subjects in each group with the given gene combination were assessed and likelihood ratios (LR) were calculated using logistic regression to combine the results of multiple genes.
Results: Early-onset CAD patients had increased, but non-significant, frequencies of PPARgamma2 Pro12/Pro12 (P = 0.39) and ENOS T-786C (P = 0.72), while BChE-K was only significantly higher in early-onset CAD patients compared to controls (P = 0.03). There were significantly more APOE epsilon4 alleles alone (P = 0.02) or in combination with BChE-K (P = 0.02) among early-onset CAD patients compared to late-onset CAD ones or controls. When combined, there was a higher prevalence of all four alleles in early-onset CAD (early-onset CAD patients: 10.7%, late-onset CAD patients: 3.3% and controls: 2.7%, P = 0.01). LR for early-onset CAD for a single allele was relatively small (1.08 for PPARgamma2 to 1.70 for APOE epsilon4). This increased to 2.78 (1.44-5.37) when combining all four alleles, therefore increasing the pre-test probability of CAD from 5% to a post-test probability of 12.7%.
Conclusions: While any single mutation causes only a mildly increased LR (none > 1.7), in combination, the likelihood of early-onset CAD increased to 2.78 with four mutations. The genetics of early-onset CAD appear to be multifactorial, requiring polygenic models to elucidate risk.