In recent years, coronary heart disease (CHD) has become the main cause of the incidence rate and mortality rate in developed China and developing countries [1]. In recent years, the incidence and mortality rate of CHD in China has increased rapidly, and the age of onset is younger. In young people (below 40 years old), about 700,000 people die from CHD every year [2, 3]. The etiology of CHD is very complex, and risk of CHD is not only affected by conventional environmental factors, but also by genetic variation [4]. Previous study has reported many environmental risk factors related to coronary heart disease, including blood lipid concentration, blood pressure, smoking, diabetes and so on [5]. In addition, inflammation also was reported associations with atherosclerosis development, which was an important risk factor for CHD [6].

Previous studies have reported that some inflammatory cytokine gene polymorphisms are associated with the risk of CHD, including cytokines in the IL-12 family [7, 8]. IL-35 is an anti-inflammatory cytokine and a member of IL-12 family. It is a heterodimer composed of p35 (IL-12A) and EBI3 subunits [9, 10]. Previous studies have reported that this genetic variation is related to the susceptibility of Alzheimer's disease and Graves' disease [11, 12], while EBI3 genetic variation may affect the risk of allergic rhinitis (AR) and tuberculosis in Chinese [13, 14]. Recently, only two studies [15, 16] have also studied the relationship between IL-35 genes polymorphism and CHD susceptibility. As far as we know, the relationship between IL-35 genes polymorphism and susceptibility to coronary heart disease has not been tested in the largest Han population in China. Lin et al. [16] performed a case–control study for Chinese population, but these participants were all Chinese Zhuang. In addition, CHD is a multifactorial disease caused by both genetic and environmental factors [4] and gene–environment interactions [17]. The aim of this study was to evaluate the influence of SNPs within IL-35 gene and their interaction with environment on susceptibility to CHD.

In this study, a total of 1847 participants were enrolled, including 921 patients with CHD and 926 normal controls. Table 2 showed the different demographic characteristics in both case and control group. The mean age of all participants is 60.9 ± 10.8 years old. There are no statistically significant differences between case and control group in males, BMI and age (p > 0.05). Compared with control group, the mean of C-reactive protein (CRP) levels, the rates for hypertension and T2DM of case group are significantly different (all p < 0.05). In addition, the case group had higher alcohol consumption and smoking rates than control group.

All genotypes of the four SNPs follow by HWE distribution in the control group (all p values in HWE testing are greater than 0.05). The C allele frequency of rs428253 gene is 20.5% in the normal control group and 30.2% in patients with CHD. The minor (G allele) allele frequencies of rs2243115 gene are 26.3% in patients with CHD and 17.8% in normal controls. Logistic regression analysis showed that both the C allele of rs428253 and the G allele of rs2243115 were independently related to the increased risk of CHD. The adjusted OR (95% CI) was 1.91 (1.28–2.64) and 1.80 (1.30–2.23), respectively. However, there was no significant correlation between CHD and rs4740 or rs568408 (Table 3).

Further investigation on gene–gene and gene–environment interaction was tested using the GMDR methods, and the cross-validation consistency and testing accuracy were calculated. With covariate adjustments, the combination consist of rs428253 and current smoking is the best model, the cross-validation consistency of which was 10/10, and testing balanced accuracy of which was 0.632 (p = 0.010, Table 4). The fore-mentioned results provided an evidence for interaction effects between rs428253 and current smoking. But all gene–gene interaction combinations obtained from GMDR test have no statistical significance. We also conducted stratified analysis for interaction effect between rs428253 and current smoking using logistic regression. We found that current smokers with rs428253-GC/CC genotype have the highest CHD risk, compared to never smokers with rs428253-GG genotype, OR (95% CI) = 3.04 (1.71–4.41) (Fig. 1).

The current study indicated that both the C allele of rs428253 and the G allele of rs2243115 were correlated with increased risk of CHD, but rs4740 or rs568408 were not statistically associated with CHD risk. The rs428253 belongs to EBI3 gene, which located on chromosome 19p13.3. The rs2243115 belongs to IL-12A subunit, which is located on chromosome 3q25.33. Previously, several studies have investigated the association between the two SNPs and inflammatory diseases [11, 12]. So far, there are few studies about the effect of IL-35 polymorphism on cardiovascular disease (CVD). Zhang et al. [14] confirmed that EBI3-rs428253 polymorphism is related to the chronic rhinosinusitis and allergic rhinitis risk reduction. So far, only two studies [15, 16] have confirmed the relationship between IL-35 genes and CHD risk, but the conclusions of the two studies are inconsistent. A study [15] for Mexico populations suggested a relationship of the EBI3 SNPs with IL-35 levels, and EBI3-rs428253 and IL-12a-rs2243115 gene polymorphisms play an important role in the mechanism of CHD risk reduction. However, a recent case–control study in the Chinese population [16] suggested a statistical correlation between the EBI3 rs428253 mutation genotype and the risk of CHD, but there was no statistical significance between the mutation genotype of IL-12A rs2243115 and the risk of CHD. The study also found that there was no significant difference in the level of IL-35 between different genotypes in the healthy control group. Although two case–control studies have been performed previously, but the two studies concluded inconsistent results, and just one study was performed for Chinese Zhuang population, but no study focused on Chinese Han population, which was the largest race in China. So, the different race may lead to different results. The biological mechanism of the association between IL-35 genes and CHD susceptibility is not well established, previous studies suggested that EBI3-rs428253 may be involved in the modification of LEF1 binding site [15, 21], and play an important role in granulocyte proliferation and differentiation [22]. SNP-rs428253 in EBI3 gene is related to the occurrence and development of CHD, which may be caused by regulating β—Catenin pathway and Treg pathway rather than by influencing the production of IL-35, because in the IL-35 levels was not different among different genotype of rs428253 in controls of study by Lin et al. [16]. In this study, we found that rs4740 and rs568408 was not associated with CHD risk, which was consistent with study by Lin et al. [16]. Although Posadas-Sánchez et al. [18] found that IL-35 genes rs4740 and rs568408 were not associated with CHD risk in Mexicans, but the two SNPs genotypes were correlated with IL-35 levels in healthy Mexicans.

The pathogenesis of CHD is very complex, and is not only affected by genetic factors and environmental factors independently, but also by the synergistic effect between them [23, 24]. We were all known that alcohol consumption and smoking were two main modifiable risk factors for CHD [25, 26]. In this study, the smoking and alcohol consumption rates were higher in cases than controls, which indicated that smoking and alcohol consumption were two risk factors for CHD. So, we performed a GMDR analysis for gene–smoking or alcohol consumption interaction, and the results suggested a significant interaction between rs428253 and current smoking. Current smokers with rs428253—GC/CC genotype have the highest CHD risk, compared to never smokers with rs428253—GG genotype. Environmental factors can cause phenotypic differences by influencing gene expression regulation. Therefore, the study of IL-35 gene–environment interaction is helpful for us to better understand the occurrence of CHD.

The C allele of rs428253 and the G allele of rs2243115 are correlated with increased risk of CHD. We also find a significant interaction between rs428253 and current smoking associated with CHD risk. In addition, we also found the interaction between rs428253 and current smoking, which added more detailed mechanism for relationship between gene, environmental risk factors and CHD susceptibility.