Background: Case-control studies have successfully identified many genetic associations for complex diseases but suffer from lack of reproducibility in the same population. Demonstrating weak genetic effect requires large sample sizes to minimize statistical bias. Based on a study examining 500 myocardial infarction (MI) patients and 500 controls from the genetically isolated Newfoundland population, we previously reported that thrombospondin-4 (THBS-4) 1186G>C variant associates with MI in women. To validate this sex-dependent association with the THBS-4 variant, we analyzed an additional 532 patients and 514 controls from the same population and the combined cohort consisting of 1032 patients and 1014 controls.
Methods: Genotyping of THBS-4 1186G>C was conducted using Taq Man 1186G>C (A3879P) (rs 1866389) genotyping technology on real-time polymerase chain reaction.
Results: The genotype distributions of THBS-4 1186G>C in the validation and combined cohorts were similar with those in our initial study, which supports genetic homogeneity in the studied population. The association of the CC genotype with MI in women (odds ratio [OR], 2.96; P = .008) reported in our initial cohort failed to achieve statistical significance in our validation cohort (OR, 1.53; P = .307) but was confirmed in the combined cohort (OR, 2.14; P = .009). In contrast to the results from the initial cohort was a significant association of the CC genotype with later onset MI in the validation (OR, 2.37; P = .029) and combined cohorts (OR, 2.22; P = .011). Moreover, the larger studied population gave statistical power to associate the CC genotype with risk of MI in the total patient population (OR, 1.58; P = .023).
Conclusion: Homozygosity for the THBS-4 1186C variant is a weak risk factor for MI especially in older women.