Association of the Apolipoprotein A-I gene polymorphisms with cardiovascular disease risk factors and atherogenic indices in patients from assam, Northeast India

Cardiovascular diseases (CVDs) are the leading cause of mortality and morbidity worldwide. In 2012, an estimated 17.5 million people died of CVDs, representing about 31.0% of all global deaths [1]. Atherosclerosis, which refers to the progressive hardening of arteries and narrowing of lumen due to fibro-fatty changes in the vessel wall, is the major pathophysiological process underlying CVDs [2]. The risk factors for atherosclerosis and CVDs are many, such as altered serum lipid levels (also called dyslipidemia), smoking, alcohol, obesity, male gender, advancing age, ethnicity and genetic predisposition, diabetes and hypertension [1-3]. The apolipoprotein A1 gene (APOA1), which is a part of the APOA1-CIII-AIV gene cluster on chromosome 11, is a major site controlling the expression of lipids and lipoproteins [4-7]. Thus, allelic variants of APOA1 that influence CVD risk are of interest. The role of two single nucleotide polymorphisms (SNPs), located in the regulatory regions of APOA1, namely G-75A and C+83T, is being increasingly appreciated in this regard. These SNPs have been implicated in various clinical conditions [8-11], but most notably with reference to CVDs and their risk factors [12-18]. The G-75A site lies in the promoter region, while the C+83T site lies in the first intron of APOA1 [19,20]. These polymorphic sites are also cleavage sites for the restriction endonuclease, MspI and thus are amenable to detection by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) [15,16,21]. These SNPs are thought to influence several quantitative traits that are risk factors for CVD, such as: serum lipids [viz. total cholesterol (TC), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), very low density lipoprotein cholesterol (VLDL-C) and triglycerides (TG)], indices of obesity [viz. waist circumference (WC) and body mass index (BMI)], and blood pressure (BP) [12-19].

Recently, the G-75A and C+83T polymorphisms were described in the native population of Assam, where significant differences were noted in their distribution as compared to populations of neighboring regions [21]. Assam is the most populated state in northeast India, an ethnically distinct region that is often regarded as the ethnological transition zone between the Indian subcontinent and east/southeast Asia. Decreased level of HDL-C is a major risk factor for CVD [1-3] and a common form of dyslipidemia in Assam [22,23]. Moreover, APOA1 codes for the major part of the proteome of HDL particles [19]. Assam also has a sizeable burden of many other conditions (high BP, diabetes, schizophrenia, increased TG, etc.) where APOA1 is implicated [23-26]. Thus, it is of interest to investigate whether the G-75A and C+83T polymorphisms have a role to play in relation to these conditions in the population of Assam.

On this basis, we decided to conduct a pilot study to explore if G-75A and C+83T SNPs of APOA1 were associated with cardiovascular risk factors in the population of Assam. Our primary objective was to determine whether G-75A and C+83T polymorphisms were associated with decreased HDL-C. Our secondary objective was to assess if these polymorphisms were associated with other CVD risk factors, viz. lipid fractions (TC, LDL-C, VLDL-C, TG); obesity indices (WC and BMI); and BP (systolic and diastolic). Additionally, we studied the associations of these SNPs with several atherogenic indices [also known as lipid ratios, namely: Castelli’s Risk Index-I (CRI-I), Castelli’s Risk Index-II (CRI-II), atherogenic index of plasma (AIP), atherogenic coefficient (AC), non-HDL cholesterol fraction (non-HDL-C)]. Atherogenic indices reflect the propensity of the serum lipids to induce atherosclerotic changes. They are increasingly recognized as valuable predictors of CVD risk [27-29], often better than the traditional lipid measures, but genetic factors influencing them have been scarcely investigated.

The baseline characteristics of the study subjects are presented in Table 1. The individuals in the case and the control groups were comparable (p >0.05) in terms of age and sex composition, and smoking and alcohol use status. Concentration of HDL-C was significantly lower in the case group than in the control group, as expected. Furthermore, the case group individuals had significantly higher (p <0.05) measures of obesity (BMI and WC), systolic and diastolic BP, lipid profile parameters (TC, TG, LDL-C and VLDL-C) and atherogenic indices in comparison to the control group individuals. None of the subjects in the study were vegetarians. During genotyping, we detected GG, GA and AA genotypes for the G-75A site, and CC and CT genotypes for the C+83T site (Figure 1).

Figure 1

Table 1

Baseline characteristics of the study subjects. (Values are expressed as mean ± SD or n (%). Categorical variables were compared by the χ² test. Continuous variables were compared by unpaired t-test.)

We explored the associations of G-75A and C+83T SNPs in the study population with cardiovascular risk factors. Although the G-75A and C+83T polymorphic sites were reported to be associated with HDL-C levels in some studies [9,15-17], we observed no such association in our study. These polymorphisms did not confer greater risk of developing decreased HDL-C dyslipidemia in the current population. Lack of association between these SNPs and low HDL-C levels was also reported by other investigators [7,12,38]. We verified our findings in different genetic models (additive, dominant, recessive and allelic) and no relationship was found. It is possible that changes in HDL-C values across the genotypic variants of G-75A and C+83T loci reported in some of the earlier studies were actually modulated by confounding factors. This is substantiated by the fact that the effects of G-75A and C+83T SNPs on phenotypic characteristics have been found to be influenced by gender differences [13,39], smoking [39] and alcohol use [40]. Such biases have not been controlled for in many of the previous studies [9,16,17]. We took care to ensure that the effects of these SNPs were assessed after controlling for potential confounders in the data analysis. Moreover, many conditions that may spuriously alter lipid levels in the cases were excluded in the design stage itself.

Among the other serum lipids, we found that the minor A allele in the G-75A locus was associated with adverse LDL-C levels. The AA and GA genotypes had high LDL-C values. This was consistent with the findings of previous investigators [4,38]. However, few studies have found the A allele to produce beneficial effects on LDL-C [41]. These discordant results may be due to linkage disequilibrium between the G-75A SNP and some other regulatory elements in close proximity that actually influence the levels of LDL-C. For example, strong linkage was observed between the A allele (of the G-75A site) and X2 allele (of the XmnI RFLP site) in the 5’-flanking region of APOA1 [4]. The X2 allele of the XmnI polymorphism has been found to be associated with LDL-C levels [42]. Future studies using sequencing would be of help in verifying such possibilities.

On the other hand, we found that the C>T transition at C+83T locus favorably affected TG and VLDL-C concentrations. Our results were in agreement to those by Ma et al. [43] who also found that the CT heterozygotes had lower TG than the CC homozygotes. The association with VLDL-C levels were expected because TG and VLDL-C metabolism are intimately related, and their values are proportional to one another [33]. It was suggested that the C+83T locus may have a role in regulating TG levels in Oriental subjects [43]. The biological mechanism causing these effects is uncertain. It was hypothesized that APOA1 variants may affect the lipase-mediated breakdown of TG particles in the plasma, thus modulating their concentration [14].

The association of the two SNPs with several atherogenic indices was a novel and significant finding of the current study. Genetic factors affecting atherogenic indices have been investigated infrequently. We found that the minor A allele at the G-75A site was associated with adverse values of CRI-I, CRI-II, non-HDL-C and AC, whereas, the minor T allele at the C+83T site was associated with favorable values of AIP. These pro-atherogenic effects of the G-75A polymorphism and the anti-atherogenic effects of the C+83T polymorphism on the atherogenic indices may be important modulators of CVD risks that are often described in relation to these two SNPs. Recent data have shown that atherogenic indices (viz. CRI-I, CRI-II, AIP, non-HDL-C, AC) are valuable predictors of CVD risk [27-30]. The usefulness of the traditional serum lipid measures (viz. TC, HDL-C, LDL-C, VLDL-C and TG) is mainly limited to predicting risk for those at the lower and the higher end of the CVD risk spectrum. In contrast, the atherogenic indices, being composite measures, take into account several lipid fractions. Thus, they reflect the bi-directional cholesterol traffic (in and outward) through the arterial intima and are more robust in predicting CVDs than individual lipid fractions [28,29].

There are contradictory reports about the effects of G-75A and C+83T SNPs on indices of obesity and BP. These SNPs were not associated with WC, BMI, systolic and diastolic BP in the current population. In conformity to the present findings, no association of the G-75A locus was detected with obesity and BP by Ma et al. [43] and de Franca et al. [6]. In contrast, Kim and Hong [18] reported association of the G-75A locus with obesity indices. On the other hand, Coban et al. [13] detected significant associations of the G-75A locus with BP, but not with obesity (BMI ≥30.0 kg/m²). For the C+83T locus, de Franca et al. [6] found no significant variation in obesity measures across the genotypes. This was in agreement with our findings. But Ma et al. [43] found the C+83T site to be associated with BP and obesity measures in diabetic subjects. In view of the above findings, it seems that these effects, if at all present, are remarkable only in some populations and that they are of an indirect nature, perhaps as a result of interaction with environmental and other genetic factors [43].

A limitation of our study was its design with respect to the secondary objectives. As our primary objective was to investigate the association of the polymorphisms with the risk of developing decreased HDL-C, we therefore had to recruit the subjects as case and control groups on the basis of HDL-C status. Thus, we could not use contingency tables and risk estimates (e.g., OR) to study the secondary objectives (i.e., association of the polymorphisms with other CVD risk factors such as TC, LDL-C, TG, VLDL-C, obesity indices, BP and atherogenic indices). However, this was circumvented by studying the associations using ANCOVA, and backing up the results with effect size calculations. It would also have been desirable to replicate and validate our findings in additional independent samples from other parts of northeast India and the adjoining regions.

The results from our pilot study suggested that the G-75A and C+83T polymorphisms of APOA1 were important modulators of CVD risk in the population of Assam. In general, the A allele at G-75A was pro-atherogenic and the C allele at the C+83T locus was anti-atherogenic. The G-75A locus was associated with LDL-C levels, while the C+83T locus was associated with TG and VLDL-C levels. But, these sites were not associated with the HDL-C status. Our study supports the hypothesis that the two SNPs do not play a major role in influencing HDL-C levels. Associations with TC, obesity and BP were also absent. Notably, these two loci were significantly associated with several atherogenic indices that are strong predictors of CVDs. Although drawn on a sample of modest size, yet these remarkable associations (with the detected effect sizes and available statistical power) from our exploratory study strongly suggested that the effect of genetic factors on atherogenic indices may be important while characterizing CVD risk. In the current population, the G-75A polymorphism seemed to be particularly important in this regard. Similar studies in other populations and larger sample groups would help in clarifying such utility of atherogenic indices.