Meta-Analysis of Apolipoprotein E Gene Polymorphism and Susceptibility of Myocardial Infarction

A number of case-control studies have been conducted to clarify the association between ApoE polymorphisms and myocardial infarction (MI); however, the results are inconsistent. This meta-analysis was performed to clarify this issue using all the available evidence. Searching in PubMed retrieved all eligible articles. A total of 33 studies were included in this meta-analysis, including 18752 MI cases and 18963 controls. The pooled analysis based on all included studies showed that the MI patients had a decreased frequency of the ε2 allele (OR = 0.78, 95% CI = 0.70–0.87) and an increased frequency of the ε4 allele (OR = 1.15, 95% CI = 1.10–1.20); The results also showed a decreased susceptibility of MI in the ε2ε3 vs. ε3ε3 analysis (OR = 0.79, 95% CI = 0.68–0.90) and in the ε2 vs. ε3 analysis (OR = 0.78, 95% CI = 0.69–0.89), an increased susceptibility of MI in the ε3ε4 vs. ε3ε3 analysis (OR = 1.26, 95% CI = 1.12–1.41), in the ε4 vs. ε3 analysis (OR = 1.22, 95% CI = 1.12–1.32) and in the ε4ε4 vs. ε3ε3 analysis (OR = 1.59, 95% CI = 1.15–2.19). However, there were no significant associations among polymorphisms and MI for the following genetic models: frequency of the ε3 allele (OR = 0.99, 95% CI = 0.96–1.02); ε2ε2 vs. ε3ε3 analysis (OR = 0.73, 95% CI = 0.40–1.32); or ε2ε4 vs. ε3ε3 analysis (OR = 1.10, 95% CI = 0.99–1.21). Our results suggested that the ε4 allele of ApoE is a risk factor for the development of MI and the ε2 allele of ApoE is a protective factor in the development of MI.


Introduction
Myocardial infarction (MI) is a leading cause of death worldwide, and is a multifactorial disease, influenced by genetic and environmental factors [1]. The main risk factors for MI include hypertension, hypercholesterolemia, diabetes, obesity, and smoking. In addition, recent studies have also shown the importance of genetic factors caused by polymorphisms in the pathogenesis of MI [2][3][4][5][6][7].
Apolipoprotein E (Apo E) is a serum glycoprotein found in circulating chylomicrons (remnants), very low density lipoproteins, intermediate density lipoproteins and high-density lipoproteins [8]. ApoE is considered as an excellent candidate gene for studying the susceptibility to coronary heart disease (CHD) and MI because of its pivotal roles in the metabolisms of cholesterol and triglyceride [9]. The most extensively studied polymorphism in the ApoE gene codes for three variant alleles: e2, e3 and e4, which yield six possible genotypes: e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4 in general population [10]. The products of the three alleles differ in their properties such as their affinity for binding low density lipoprotein receptors and lipoprotein particles; therefore, this ApoE polymorphism could affect the serum levels of cholesterol and triglyceride, thus contributing to the progression of atherosclerosis. In fact, ApoE polymorphisms have been found to be associated with many lipid-related diseases and cardiovascular and cerebrovascular diseases [11][12][13][14].
Numerous studies have been conducted to explore the association of this ApoE polymorphism and CHD; some of the studies found a significant association between the ApoE e4 allele and CHD [15][16][17]. A meta-analysis conducted in 2004 provided evidence that the e4 allele of ApoE was a risk factor for the development of CHD [18]. Another meta-analysis conducted in 2013 further confirmed this finding in a Chinese population [19]. However, no meta-analysis has been conducted to explore the association between this ApoE gene polymorphism and MI. In spite of the presence of advanced CHD, only a subset of patients develops MI during their life. The reasons for these individual differences in susceptibility to MI are poorly understood. Therefore, it is important to explore the association between ApoE gene polymorphisms and MI. In fact, a number of casecontrol studies have been conducted to clarify the association between ApoE gene polymorphisms and MI ; however, the results are inconsistent. Therefore, we conducted this metaanalysis including all of the evidence produced to date to explore this issue.

Search strategy
We searched all published studies in the Pubmed database (up to January 20, 2014) using the following combination of keywords: ''Apolipoprotein E'' OR ''ApoE'' AND ''acute coronary syndrome'' OR ''myocardial infarction'' AND ''polymorphism'' OR ''polymorphisms'' OR ''variants'' OR ''variant''. In addition, manual searches for related articles were also performed to avoid missing any relevant studies.

Inclusion and exclusion criteria
The inclusion criteria for identified articles were as follows: 1) Case-control studies with full text articles on the relationship of ApoE polymorphisms and MI; 2) sufficient data for estimating an odds ratio (OR) with 95% confidence interval (CI). Those not designed as case-control studies, systemic reviews, those not written in English or Chinese, and those that provided no usable data, were excluded.

Data extraction
Two authors independently extracted the data from all included studies using a predesigned data extraction table. The following information was extracted from each included article: first author, year of publication, ethnicity and country, source of controls, total numbers of MI cases and controls, distribution of genotypes and alleles in MI cases and controls, and evidence of conforming to the Hardy-Weinberg equilibrium (HWE).

Statistical analysis
We firstly used chi-squared (x 2 ) test and and I 2 statistic to assess heterogeneity across studies. A fixed effect model (Mantel-Haenszel) was used in the absence of heterogeneity. Otherwise, the random effect model (DerSimonian-Laird) was adopted. The strength of the association between the ApoE gene polymorphism and MI was assessed by odds ratios (ORs) with the corresponding 95% CI for each study. The ORs and their 95% CIs were assessed for the following seven genetic models: 1) e2e2 vs. e3e3; 2) e2e3 vs. e3e3; 3) e2e4 vs. e3e3; 4) e3e4 vs. e3e3; 5) e4e4 vs. e3e3; 6) e2 vs. e3; 7) e4 vs. e3. The allele frequencies of e2, e3 and e4 were also assessed using the same method. Cumulative meta-analysis was also performed for the above genetic models. Subgroup analysis for ethnicity (Asian and Caucasian) was also performed. To find potential outliers, influence analysis was performed by omitting each study in turn. A funnel plot, calculated using Begg's and Egger's tests, was adopted for assessing potential publication bias. Statistical analysis was conducted using STATA statistical software (version 11; StataCorp, College Station, Texas, USA). A P value less than 0.05 was considered statistically significant.

Literature selection and study characteristics
One hundred and thirty two articles were retrieved from PubMed, 79 of which were excluded after screening the titles and abstracts (58 were irrelevant studies, 13 were reviews and eight were not published in English or Chinese). Fifty-three articles were selected for detailed assessment, which excluded a further 20 articles (seven were not case-control studies, eight had no usable data (no case and control numbers according to the genotypes) and five were not about MI). Finally, 33 studies were included in this meta-analysis, which included 18752 MI cases and 18963 controls. The detailed selection procedure is shown in Figure 1. There were three studies did not follow the HWE. The detailed characteristics of the included studies are shown in Table 1. The present study met the PRISMA statement requirements (Checklist S1 and Figure 1).  Quantitative data synthesis The meta-analysis of the included studies showed that there was significant association between the ApoE gene polymorphism and MI. The results showed that the MI patients had a decreased frequency of the e2 allele (OR = 0.78, 95% CI = 0.70-0.87, Figure 2) and an increased frequency of the e4 allele (OR = 1.15, 95% CI = 1.10-1.20, Figure 3). The results also showed a decreased susceptibility of MI in the e2e3 vs. e3e3 analysis (OR = 0.79, 95% CI = 0.68-0.90, Figure S1), and in the e2 vs. e3 analysis (OR = 0.78, 95% CI = 0.69-0.89, Figure S4), and an increased susceptibility of MI in the e3e4 vs. e3e3 analysis (OR = 1.26, 95% CI = 1.12-1.41, Figure S2) in the e4e4 vs. e3e3 analysis (OR = 1.59, 95% CI = 1.15-2.19, Figure S3) and in the e4 vs. e3 analysis (OR = 1.22, 95% CI = 1.12-1.32, Figure S5). However, there were no significant associations among polymorphisms and MI for the following genetic models: frequency of e3 allele (OR = 0.99, 95% CI = 0.96-1.02); e2e2 vs. e3e3 analysis (OR = 0.73, 95% CI = 0.40-1.32); e2e4 vs. e3e3 analysis (OR = 1.10, 95% CI = 0.99-1.21). The detailed results are shown in Table 2. Cumulative analysis further confirmed the results ( Figure 4 and Figure S6).

Sensitivity analysis
We conducted influence analysis to assess the sensitivity of each individual study on the pooled ORs by sequential omission of each individual study. The results suggested that no individual study significantly affected the pooled ORs in the e2 allele and e4 allele frequency analysis (Figure 5), and in the e2e3 vs. e3e3 analysis, e3e4 vs. e3e3 analysis and e4e4 vs. e3e3 analysis ( Figure S7).

Publication bias
Funnel plots examined potential publication bias qualitatively and no obvious asymmetry was observed in any genetic model, as shown in Figure 6. Furthermore, the results from Begg's and Egger's tests did not provide any evidence of publication bias (Table S1).

Discussion
To the best of our knowledge, this is the first meta-analysis to evaluate the association between an ApoE polymorphism and susceptibility of MI. In this meta-analysis, we discovered an increased susceptibility of MI in the e4 allele frequency analysis. Moreover, the individuals with e2e4 genotype, e3e4 genotype and e4e4 genotype had a significantly higher susceptibility of developing MI compared to those with the e3e3 genotype. Therefore, it is reasonable to assume that the e4 allele of ApoE is an risk factor for the development of MI. These results were consistent with a previous meta-analysis, which showed that e4 allele of ApoE is a risk factor for the development of CHD [11,12].
In addition, we found a decreased susceptibility of MI in the e2 allele frequency analysis and in the e2e3 vs. e3e3 analysis, which indicate that the e2 allele is a protective factor in the development of MI. Cumulative meta-analysis also confirmed these findings.
Considering the large sample size in the pooled analysis in this meta-analysis, we believe that our results are robust and reliable.
ApoE is a multifunctional protein that plays an important role in the metabolism of cholesterol and triglycerides, by binding to its receptors to help mediate clearance of chylomicron and remnant particles [53]. The three common isoforms, e2, e3 and e4, have different receptor-binding abilities and could yield different circulating levels of cholesterol and triglycerides. Compared with e3 homozygotes, carriers of the e2 allele have lower circulating cholesterol levels, whereas carriers of the e4 allele appear to have higher plasma levels of total and low-density lipoprotein cholesterol [54]. According to these mechanisms, our meta-analysis suggested that carrying the e4 allele is a risk factor for MI and that the e2 allele has a protective role in the development of MI. When stratifying the studies by ethnicity, the e4 allele remained a risk factor and the e2 allele was still protective in the development of MI among Caucasian populations; however, only the e4 allele remained as a risk factor for MI among Asian population. This may be due to the small sample size in the analysis among Asian populations; in fact, there were only four studies that included Asian populations [19,20,26,36]. Therefore, further studies are warranted among Asian populations. In addition, genotype distributions in the controls from Scaglione's study [28], Bustan's study [49] and Zende's study [52] were not in agreement with HWE, therefore, the results may be biased. However, sensitivity analysis suggested that the pooled results were not significantly changed after excluding the three studies (data not shown). This may be due to the large sample size even though the three studies were excluded.
Although the primary results of this meta-analysis are suggestive, some limitations still exist. First, between-study heterogeneity existed in some of the genetic model analysis, which may have affected the results of the present meta-analysis, although a random effects model was adopted for these analyses. Second, publication bias may have occurred because our analyses were based wholly on published studies only in English and Chinese. Third, the results of this meta-analysis were based on unadjusted estimates because of the lack of adjusted estimates. Currently, some risk factors have been identified for MI, such as hypertension, hypercholesterolemia, diabetes, obesity and smoking. A more precise analysis should be performed if these data could be extracted from primary articles.
In conclusion, this comprehensive meta-analysis has evaluated all published data currently available on the association between the ApoE polymorphism and MI. Our meta-analysis suggested that the e4 allele of ApoE is an risk factor for the development of MI and the e2 allele of ApoE is a protective factor in the development of MI. This may be explained by the fact that e4    Checklist S1 PRISMA Checklist. (DOC)