Our study demonstrated an association between Gly1057Asp polymorphism in IRS-2 and CAD. Additionally, we produced evidence that the Gly1057Asp polymorphism in IRS-2 interacts with DM in relation to risk of CAD; there is a synergistic effect toward CAD between pathogenicity of DM and that of the Gly allele.
In our study population, the Gly/Gly and Gly/Asp genotype frequencies for CAD patients were significantly greater than in control subjects (Table 2). These results suggest that the Gly allele positively correlates with the occurrence of CAD, and therefore the Asp allele has a protective effect against CAD. In multivariate binary logistic regression, the odds ratio of CAD was 2.008 for patients with Gly/Gly and Gly/Asp genotypes and 1.847 for those with DM. These results suggest that the Gly1057Asp polymorphism in IRS-2 is independent of DM as a risk factor for CAD and that both Gly allele at codon 1057 in IRS-2 and DM are potent risk factors for CAD (Table 3). Importantly, we found that the concurrence of Gly allele with DM is associated with an increased susceptibility to CAD (Table 4). In patients with either Asp/Asp genotypes or Gly/Gly and Gly/Asp genotypes, the odds ratio for CAD in diabetics is almost twice that of their non-diabetic counterparts, which is compatible with the results reported in the Framingham study . In non-diabetics, the odds ratio for CAD in those with Gly/Gly and Gly/Asp genotypes was 1.922, which was about twice the ratio for those with Asp/Asp genotypes. Similarly, in diabetics, the odds ratio for CAD in those with Gly/Gly and Gly/Asp genotypes was 3.629, which was roughly twice the ratio for patients with Asp/Asp genotypes. Further, the odds ratio for CAD in diabetics with Gly/Gly and Gly/Asp genotypes was 3.629, which was roughly four times the ratio found in non-diabetics with Asp/Asp genotypes (Table 4). These results suggest that the pathogenic ability of the Gly allele to cause CAD is similar to that of DM. They also suggest a synergistic effect between the pathogenicity of DM and that of the Gly allele. We found that there was no correlation between Gly1057Asp polymorphism in IRS-2 and insulin resistance irrespective of BMI values (Table 5) and obesity (Additional file 4). These results probably exclude the possibility that the Gly1057Asp polymorphism in IRS-2 contributes to CAD through altering insulin resistance.
Several studies have already reported that genetic variants caused by a SNP are associated with CAD, including adiponectin receptor 2 gene [29, 30], apolipoprotein E gene , and genes related to regulation of blood lipid levels . In current study, we found the concurrence of Gly1057Asp polymorphism in IRS-2 with DM affects the susceptibility to CAD. This interaction between Gly1057Asp polymorphism in IRS-2 and DM could be a kind of gene-environmental or gene-gene interaction. DM itself or genes contributory to development of DM could interact Gly1057Asp polymorphism in IRS-2 to influence the susceptibility to CAD. Similar findings for interrelation of genetic variants and DM in relation to disease traits have been reported. Genetic variants of transcriptional factor 7-like 2 gene are associated with CAD, which are significantly modulated by the presence of type 2 DM . Besides, the +183 A/G polymorphism at the 3’-untranslated region of interleukin (IL)-18 genes is associated with the circulating IL-18 levels, especially apparent in patients with type 2 DM and metabolic syndrome . The frequency of Gly allele was 58.6% in our study subjects, which was approximately equal to that found in Chinese and Finnish patients . We found that Gly1057Asp polymorphism in IRS-2 was not associated with insulin resistance, which was consistent with studies of Finnish subjects . However, the Asp allele was reported to be correlated with better insulin sensitivity in specific groups in Danish  and Italian  populations. This discrepancy may be due to differences in ethnics, or to heterogeneous demographic characteristics. Thus, further investigation is needed to clarify the relationship between Gly1057Asp polymorphism in IRS-2 and insulin resistance.
DM is a potent risk factor for CAD . However, the susceptibility to CAD is different among diabetic patients. In diabetic patients concomitantly with other conventional risk factors for CAD such as hypercholesterolemia and hypertension, the risk for cardiovascular events in the following 5 to 10 years actually increases much more than that for their counterparts [35–37]. Therefore, genetic screening for Gly1057Asp polymorphism in IRS-2 may be a plausible method to early detect diabetic patients with high susceptibilities to CAD.
Our study is a cross-sectional study. The inherent weakness of cross-sectional studies is an inability to indicate causality because the predictor variable is not shown to precede the outcome. However, the predictor variable that we were interested in for this study, Gly1057Asp polymorphism in IRS-2, was determined by the time the study subjects were born. Hence, the results of our study do demonstrate a causal relationship between Gly1057Asp polymorphism in IRS-2 and CAD. Nonetheless, there were several limitations to this study. First, the number of patients for our main objective, investigating the relationship between Gly1057Asp polymorphism in IRS-2 and CAD, was limited. The number of control patients was 170, which is less than the calculated sample size of 200 patients required to achieve a power of 80% with a two-tailed α of 0.05. Thus, a larger scale study is needed to strengthen the present results. The study was also limited in that it was a hospital-based study. There may have been selection bias. Our study subjects were not truly representative of the overall population. Therefore, a general population-based study is needed to confirm our findings. Using multi-slice computed tomography to screen coronary arteries may be a feasible method for this purpose. When we investigated the relationship between Gly1057Asp polymorphism in IRS-2 and insulin resistance we studied only 190 of the total 454 patients. Also, we checked the insulin resistance by calculating the HOMA-IR index rather than by using the gold standard method euglycemic hyperinsulinemic clamp. Hence, our conclusion that there is no correlation between Gly1057Asp polymorphism in IRS-2 and insulin resistance should be further confirmed by an investigation using euglycemic hyperinsulinemic clamp on a larger scale. Finally, although we showed a correlation between Gly1057Asp polymorphism in IRS-2 and CAD, we did not afford the molecular mechanisms underlying this correlation.