Genetic Variants Associated With Myocardial Infarction Risk Factors in Over 8000 Individuals From Five Ethnic GroupsCLINICAL PERSPECTIVE
The INTERHEART Genetics Study
Background— Myocardial infarction (MI) is a leading cause of death globally, but specific genetic variants that influence MI and MI risk factors have not been assessed on a global basis.
Methods and Results— We included 8795 individuals of European, South Asian, Arab, Iranian, and Nepalese origin from the INTERHEART case-control study that genotyped 1536 single-nucleotide polymorphisms (SNPs) from 103 genes. One hundred and two SNPs were nominally associated with MI, but the statistical significance did not remain after adjustment for multiple testing. A subset of 940 SNPs from 69 genes were tested against MI risk factors. One hundred and sixty-three SNPs were nominally associated with a MI risk factor and 13 remained significant after adjusting for multiple testing. Of these 13, 11 were associated with apolipoprotein (Apo) B/A1 levels: 8 SNPs from 3 genes were associated with Apo B, and 3 cholesteryl ester transfer protein SNPs were associated with Apo A1. Seven of 8 of the SNPs associated with Apo B levels were nominally associated with MI (P<0.05), whereas none of the 3 cholesteryl ester transfer protein SNPs were associated with MI (P≥0.17). Of the 3 SNPs most significantly associated with MI, rs7412, which defines the Apo E2 isoform, was associated with both a lower Apo B/A1 ratio (P=1.0×10−7) and lower MI risk (P=0.0004). Two low-density lipoprotein receptor variants, 1 intronic (rs6511720) and 1 in the 3′ untranslated region (rs1433099) were both associated with a lower Apo B/A1 ratio (P<1.0×10−5) and a lower risk of MI (P=0.004 and P=0.003, respectively).
Conclusions— Thirteen common SNPs were associated with MI risk factors. Importantly, SNPs associated with Apo B levels were associated with MI, whereas SNPs associated with Apo A1 levels were not. The Apo E isoform, and 2 common low-density lipoprotein receptor variants (rs1433099 and rs6511720) influence MI risk in this multiethnic sample.
Received August 11, 2008; accepted December 4, 2008.
The World Health Organization estimates that by the year 2010 coronary heart disease (CHD) will be the leading cause of death globally with about 80% of the mortality occurring in developing countries.1,2 Recently, the INTERHEART case-control study showed that 9 risk factors (dyslipidemia, diabetes, hypertension, abdominal obesity, tobacco exposure, physical inactivity, psychosocial stressors, low fruit and vegetable intake, and no alcohol consumption) contribute to the risk for myocardial infarction (MI) globally.3 The ratio of apolipoprotein (Apo) B/A1 was the strongest MI risk factor and accounted for 54% of the total population attributable risk (PAR) on MI.4 These 9 risk factors are themselves potentially influenced by genetic variants, which could act on their own, or in combination with other genetic or lifestyle factors. Genetic variants that influence these risk factors may also be associated with MI.5,6 Recent studies that have identified genetic associations with MI and MI risk factors have largely been conducted among whites.5,7 However, whether these findings can be extended to other ethnic groups remain to be demonstrated. In fact, the recent association between chromosome 9 variants and CHD, which was observed in whites, was not replicated in African-Americans.8 In addition, single-nucleotide polymorphisms (SNPs) in some lipid-related genes seem to be strongly associated with plasma lipids5,9,10 but inconsistently with heart disease.5,10 To address these questions, we performed a study of 103 genes known or suspected to influence MI or MI risk factors (intermediate phenotypes) in 5 ethnic groups from the INTERHEART study.
Editorial see p 3
Clinical Perspective see p 16
INTERHEART was a standardized case-control study of acute MI from 262 centers in 52 countries. Twelve thousands four hundred sixty-one cases and 14 637 controls were enrolled between February 1999 and March 2003.3 Women and men were recruited in Asia, Europe, the Middle East, Africa, Australia, North America, and South America. The study protocol was approved by the ethics committees in all participating centers and all participants provided informed consent. The details of selection, exclusion, and baseline characteristics have been previously reported.3 Briefly, cases of incident acute MI, presenting to a hospital within 24 hours of symptom onset, were age- (±5 years) and sex-matched with controls who were hospital- or community-based individuals with no previous diagnosis of heart disease or history of exertional chest pain. Information about demographic factors, lifestyle (smoking, leisure time physical activity, and dietary patterns), personal and family history of cardiovascular disease, and risk factors (hypertension, diabetes, and psychosocial factors) was obtained using structured questionnaires administered by study personnel. Description of the measurement techniques have been previously published.3,11
Blood and DNA Samples
Nonfasting blood samples (20 mL) were drawn and centrifuged within 2 hours of admission, and frozen immediately. Blood was drawn from cases within 24 hours of symptom onset. Samples were shipped in nitrogen vapor tanks to a central blood storage site and stored in liquid nitrogen. In this study, samples from all countries were analyzed at the Clinical Trials and Clinical Research Laboratory at the Hamilton General Hospital, Hamilton Health Sciences for Apo B and Apo A1 using previously reported methods.3,4 DNA was extracted from buffy coats at the Clinical Trials and Clinical Research Laboratory and at the McGill University and Génome Québec Innovation Centre in Montreal using the same DNA extraction method (Gentra Instruments, Minneapolis, Minn).
Selection of Genes and SNPs
The selection criteria for genes and SNPs has been previously published.12 Briefly, genes were selected for this panel if (1) there was a published genetic association with MI or one of the INTERHEART risk factors or (2) they belonged to biochemical pathways involved in the etiology of MI or an intermediate phenotype. Three classes of SNPs were included (1) tagging SNPs for each gene (±10 kb of the gene to include possible regulatory regions) were selected using the software “linkage disequilibrium (LD)-select”13 on 3 HapMap populations (Chinese, Yoruban, and white),14 (2) nonsynonymous SNPs, and (3) SNPs previously associated with MI or an intermediate phenotype. These 3 classes of SNPs together yielded an average of 15 SNPs/gene. The full list of 103 genes and SNPs12 is available online: http://www.sciencedirect.com/science/MiamiMultiMediaURL/B8JDD-4R17R80-9/B8JDD-4R17R80-9-K/43612/84a226c8a405ff69f70b65774ee93604/f.rtf. One thousand eighty-six SNPs from 69 genes were known or suspected from the literature to be associated with at least 1 of 8 INTERHEART risk factors (Apo B/A1, hypertension, diabetes, abdominal obesity, tobacco use, physical inactivity, psychosocial stress, or alcohol use). No genes putatively associated with fruit and vegetable intake were added to the gene panel, and therefore no testing against this intermediate phenotype was performed.
Genotypes were produced using the Illumina GoldenGate technology15 and the BeadStudio software package. Some SNPs (n=61) were genotyped on the Sequenom platform (iPLEX Gold Assay, Sequenom, Cambridge, Mass) described in Ehrich et al.16 Successful assays (genotypes called for >90% of samples) were generated for >95% of the SNPs attempted on Illumina and for >84% of the SNPs attempted on Sequenom. The reproducibility was 99.9% for the Illumina panel and 99.8% for the Sequenom panel. Of the 1536 SNPs on the panel, 14 failed our quality control tests. From the remaining 1522 SNPs, 162 were excluded because they were not polymorphic in all ethnic groups, and 36 SNPs did not pass Hardy Weinberg equilibrium using the permutation version of the exact test after 10 000 permutations.17 A SNP was rejected if it did not pass Hardy Weinberg equilibrium in the controls of any 1 of the 5 ethnic groups (P<0.001). This yielded a total of 1324 SNPs that were tested directly against MI, and 940 SNPs that were tested against at least 1 of 8 intermediate phenotypes (Table 1). Blood samples were available for 21 508 (79%) of the 27 098 INTERHEART cases and controls (16 353 men and 5155 women). We selected the 3 ethnicities with the greatest amount of available DNA to be genotyped: European, South Asian, and Arab (9751 cases and controls).
Quality Control, Assessment of Population Structure, and Matching
We used the program STRUCTURE18 to infer relative ancestry from genotype data, and we determined that the Nepalese (initially classified as South Asians) and Iranian (initially classified as Arabs) samples should be treated as distinct ethnic groups.14 In addition, we identified individuals for whom the genetically inferred ancestry did not correlate with self-reported ethnicity and these 104 individuals (1.2%) were excluded. After exclusion of samples with a self-reported ethnicity different from the STRUCTURE analysis and other problematic samples (duplicates, first degree relatives, etc), 8795 cases and controls were available for this study. For association tests with MI, cases and controls were matched 1:1 for sex, age (±5 years), and ethnicity. We thus excluded 20 cases where no appropriate control match existed, and 741 controls where no case matched, resulting in 8034 cases and controls from 5 ethnic groups for MI analysis (Figure 1).
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
All SNPs that were polymorphic in all ethnic groups and that passed the Hardy Weinberg equilibrium test (n=1324) were tested for association with acute MI using logistic regression. Both nominal and Bonferonni corrected probability values were calculated. The number of nominal associations for 1324 SNPs expected by chance was calculated using permutations (n=1000). Linear regression (adjusting for age, sex, and ethnic group) was used to assess the association between a SNP and quantitative traits (putative intermediate phenotypes), and for categorical outcomes, the per-allele odds ratio (OR) of the minor allele was calculated using logistic regression. A Bonferroni correction for multiple testing was used to determine statistically significant associations with intermediate phenotypes. Significant associations with an intermediate phenotype were corrected for the number of SNPs tested against that intermediate phenotype (Table 1). SNPs associated with intermediate phenotypes were compared in MI cases versus controls (matched for age [±5 years], sex, and ethnicity) using conditional logistic regression under an additive model. Permutations were used to calculate the significance of SNPs versus MI. The PAR for SNPs and INTERHEART risk factors was calculated using the methods described by Benichou and Gail.19 We updated a recent meta-analysis of Apo E SNPs and MI20 by adding the INTERHEART data using a random effects model. The consistency of findings across studies was assessed using the I2 statistic.21
Characteristics of this INTERHEART Sample
The 8034 individuals (matched cases and controls) included in this study are of South Asian (n=2346), Arab (n=1498), European (n=3666), Iranian (n=402), and Nepalese (n=122) origin (Table 2). Consistent with the total INTERHEART sample,3 with the exception of alcohol intake, selected risk factors are significantly associated with MI (Table 2).
Association of SNPs With MI and Intermediate Phenotypes
When SNPs were tested directly against acute MI, 102 associations (P<0.05) were observed (Supplementary Table S1), although no significant associations with MI remained after correcting for the number of statistical tests. However, the 102 significant associations exceeded the number expected by chance (n=61) and in 1000 permutations, 102 or more significant SNPs were observed only 3 times. In addition, 163 SNPs were significantly associated (P<0.05) with an intermediate phenotype and of these, 13 remained significant after correction for multiple testing (Tables 1 and 3⇓). Eleven of these 13 SNPs were from 4 genes (Apo E, cholesteryl ester transfer protein [CETP], low-density lipoprotein receptor [LDLR], Apo B) and were associated with the Apo B/A1 ratio (P<7.5×10−5) (Table 3). Collectively these 11 SNPs accounted for 4.3% of the variance of Apo B/A1 levels. Eight of the 11 SNPs associated with the Apo B/A1 ratio were significantly associated with Apo B (in the genes Apo E, LDLR, and Apo B). The CETP SNPs that are associated with the Apo B/A1 ratio were in high LD with each other (r2>0.91) and were significantly associated with ApoA1. Furthermore, 7 of the 8 SNPs associated with Apo B were nominally associated with MI (P<0.05) and the eighth one almost reached significance (P=0.06). In contrast, the CETP SNPs that were significantly associated with Apo A1 were not associated with MI (P=0.17 to 0.88) (Table 3).
We also observed that 2 SNPs (rs904096 and rs698) from the alcohol dehydrogenase 1C gene were associated with alcohol consumption after correction for 13 tests (P<0.0038). The 2 SNPs are in high LD with one another (r2>0.97 for all 5 ethnicities), and were not associated with MI (Table 1). No other SNPs were associated with intermediate phenotypes after correction for multiple testing (Table 1).
Of the 13 SNPs that were significantly associated with an intermediate phenotype and then tested for association with MI, 3 were significant at the P≤0.004 level under an additive model (Table 4). The SNP rs7412 (MAF range: 3.7% to 6.8%), which defines the ε2 isoform of the APOE gene, had an OR of 0.78 (95% CI: 0.70 to 0.89; P=0.0004). In the LDLR gene the SNP rs1433099 (MAF range: 26.5% to 34.8%), had an OR of 0.90 (95% CI: 0.84 to 0.96; P=0.002) and rs6511720 (MAF range: 2.1% to 14.1%) had an OR of 0.86 (95% CI: 0.77 to 0.95; P=0.004). These 2 SNPs were not in significant LD with each other. After adjustment for Apo B/A1 levels, the LDLR SNP rs1433099 remained independently associated with MI (P=0.0065) as well as after adjustment for all 9 INTERHEART risk factors.
The SNP rs429358 that defines the ε4 isoform of Apo E, was also associated with the Apo B/A1 ratio (P=1.0×10−10). We evaluated the association between Apo E isoforms and acute MI by combining the genotypic data from these 2 SNPs (rs7412 and rs429358). The relative frequencies of the isoform classes were consistent across ethnic groups, with the ε3 being the most common (66.7% to 79.5%), the ε4 accounting for 11.1% to 22%, and the ε2 allele being the least common (6.5% to 11.3%) (Table 5). Results by genotype are provided in Supplementary Table S2. The Apo B/A1 ratio increased from ε2 carriers to ε3 homozygotes to ε4 carriers within all major ethnic groups (Supplementary Table S3). Plasma Apo B also increased in a stepwise fashion whereas Apo A decreased. When comparing the mean Apo B/A1 ratio among control individuals (ε2=0.67, ε3=0.84, ε4=0.90), the overall P for trend was 0.0001. Apo E isoform status accounted for 3.7% of the variance of Apo B, and 2.9% of the variance of the Apo B/A1 ratio, but only 0.3% of the variance of Apo A levels. We observed an approximately linear relationship of Apo E genotypes with MI risk (Figure 2). The PAR for the Apo E isoforms considering the graded risk increase from ε2 to ε3 to E4 is 17.5%. The per genotype class change from ε2 to ε4 for MI was 1.19 (95% CI: 1.09 to 1.30; P=0.0001), reflecting the odds of MI for ε3 versus ε2 (1.29; 95% CI: 1.10 to 1.51), and for ε4 versus ε2, (1.47; 95% CI: 1.22 to 1.76). When adjusted for the Apo B/A1 ratio, the ORs became nonsignificant for both the ε3 versus ε2 (OR=1.05; 95% CI: 0.89 to 1.23) and the ε4 versus ε2 (OR=1.08; 95% CI: 0.89 to 1.30) comparisons. We updated a recent meta-analysis comparing the effect of Apo E isoforms on the risk of coronary artery disease,20 and the totality of the data supports our findings that the ε2 isoform lowers the risk of CAD, whereas the ε4 isoform increases the risk of coronary artery disease (Figures 3 and 4⇓).
The minor alleles of the two SNPs from the LDLR locus were both associated with a lower risk of MI. The SNP rs6511720 had a MAF of ≈10% and its presence was associated with a lower Apo B, a lower Apo B/A1 ratio, as well as a lower risk of MI (OR=0.86; 95% CI: 0.77 to 0.95). The PAR associated with the risk allele was 24.2%, however, adjustment for Apo B/A1 nullified the association with MI. The other SNP (rs1433099) was more common (≈30% MAF) and its presence was also associated with lower Apo B, lower Apo B/A1, and a reduced risk of MI (OR=0.90; 95% CI: 0.84 to 0.96). The association of rs1433099 remained significant after adjustment for Apo B/A1 and for all 9 INTERHEART risk factors.
The incremental predictive value of the SNP rs1433099 that remained independent of all INTERHEART risk factors was determined by first calculating the PAR of the 9 risk factors on MI (89.6%; 95% CI: 86.4 to 92.2%), and next by adding the LDLR SNP rs1433099 into the model. With this addition, the PAR increased by 1.6% so that the total PAR for the 9 risk factors and rs1433099 on MI was 91.2% (95% CI: 88.3% to 93.8%).
Recent genome-wide association studies have produced several robust genetic associations with lipid levels and CHD.5,10,22 Our results are consistent with the results of these recent genome-wide association studies of blood lipid levels as 4 of the gene loci we identified to be significantly associated with Apo B/A1 levels (Apo E, Apo B, LDLR, and CETP) were also identified by genome-wide association studies.5,10 Interestingly, though we identified 11 SNPs from 4 genes that were strongly associated with the Apo B/A1 ratio (P<10−5) only SNPs from genes associated with Apo B, are significantly associated with MI. The 3 SNPs, which were strongly associated with Apo B/A1 levels but not with MI, were from the CETP gene and had a strong effect on Apo A1.
Some previous large studies have similarly shown strong evidence of association between CETP genotypes and Apo A1 or HDL cholesterol levels, but no corresponding association with MI risk.5,23 Our collective findings, raise the question of whether HDL cholesterol as measured by current assays, is a marker or mediator of coronary artery disease. Certain single-gene conditions characterized by low or very low HDL levels have premature coronary artery disease as one of their manifestations (for example Tangier disease, which results from mutations in the ABCA1 gene), whereas others leading to similarly low HDL levels do not (for example ApoA1-Milano). In fact, recent data has shown that HDL particles containing Apo A1-Milano is more effective than HDL containing normal Apo A1 at maintaining endothelial cell homeostasis under stress, due to upregulation of endothelial nitric oxide synthase expression and downregulation of vascular cell adhesion molecule expression.24 Thus, qualitative differences in the HDL particle, mediated through genetic variability in Apo A1 and other genes, may be as important as quantitative differences in plasma HDL level in determining MI risk.
Our finding that SNPs in the alcohol dehydrogenase 1C gene were associated with alcohol consumption has been previously reported.25 In addition, variants in this gene have also been associated with alcohol dependence26 and alcoholism,25 and they are known to display large differences in allele frequency among populations.27 Genetic variants at this locus have also been identified that interact with alcohol consumption to increase heart disease.28,29 However, this finding has not been observed in all studies.30
We observed that 3 SNPs in 2 lipid metabolism genes were associated with the levels of Apo B/A1 ratio and with acute MI. The association of the Apo E isoforms with MI became nonsignificant after adjustment for the Apo B/A1 ratio. Although a similar observation was made for one of the LDLR SNPs (rs6511720), the other one, rs1433099, remained independently associated with acute MI even after adjustment for Apo B/A1, or adjustment for all 9 INTERHEART risk factors.
Previous genetic studies of the Apo E ε4 isoform have provided evidence that it is significantly associated with increasing concentrations of LDL and Apo B/A1 ratio as well as MI.23 The ε2 isoform, however, has been more controversial with some meta-analyses31 but not all20 showing an inverse association with MI. Song et al31 concluded that ε4 was associated with MI risk (OR=1.30; 95%: 1.18 to 1.44) but that ε2 did not protect against the risk of MI (OR=0.95; 95% CI: 0.84 to 1.14) when compared with the common ε3 isoform. The more recent meta-analysis by Bennet et al20 reported that ε2 carriers had a 20% lower risk of CHD (95% CI: 10% to 30%) and ε4 carriers had a 6% increased risk of CHD (95% CI: −1% to 13%), which only approached statistical significance. Our study findings agree with the largest previous study among people of European ancestry, the ISIS case-control study that included ≈10 000 cases and controls.23 We also observed that the ε4 isoform significantly increased MI risk and the ε2 isoform was significantly protective against MI. Addition of the INTERHEART data to the meta-analysis of Bennet et al indicated that ε2 carriers have a 20% (95% CI: 11% to 29%) lower risk of CHD when compared with the common ε3/ε3 genotype, and the ε4 isoform was associated with a 7% (95% CI: 1% to 13%) increase in CHD which now becomes significant (Figures 3 and 4⇑). Furthermore, our study demonstrates that Apo E has a substantial impact on MI in multiple ethnic populations; and that the impact of Apo E isoforms can be wholly explained by their impact on the Apo B/A1 ratio.
We observed that 2 common SNPs from the LDLR locus were associated with Apo B/A1 ratio and with acute MI. Rare coding mutations in the LDLR gene cause familial hypercholesterolemia,32 but these do not explain a high percentage of plasma lipoprotein variation in the general population. Recently some common SNPs in LDLR have been shown to affect lipid levels and coronary artery disease in people of European origin.5,10
Our finding that 2 common LDLR variants contribute to the risk of MI highlights the importance of variants of this gene, not just for familial hypercholesterolemia families, but for the general population in multiple ethnicities. Our results for these 2 SNPs in LDLR are likely independent and not due to LD, as the 2 SNPs had an r2 of only 0.15 to 0.62 in all 5 ethnicities. rs6511720 is in the first intron, a location where regulatory genetic elements, are commonly encountered; and rs1433099 is 32 Kb downstream in the 3′ UTR, a region known to be important in the regulation of mRNA stability in many genes. After adjustment for Apo B/A1 ratio, the association between rs6511720 and MI became null, whereas rs1433099 remained significantly associated with MI (P=0.0065). Interestingly the association of rs1433099 with MI remained after adjustment for all 9 INTERHEART risk factors (P=0.0039); hence, its association with MI could not be explained by its effect on plasma Apo B/A1 or any of the other risk factors. Recent work in the Atherosclerosis Risk In Communities study found that SNPs in the 3′ UTR, including rs1433099, were associated with LDL-cholesterol in whites, but not in African Americans.33 This study also observed functional variation between the alleles using reporter assays. Contrary to expectation, the allele that led to increased LDL-C levels also had higher mRNA stability (and would therefore be expected to result in increased, rather than decreased, expression of the LDL receptor protein). This is an interesting finding in light of our result of an association with MI that is independent of Apo B levels. However, we cannot rule out LD with another functional SNP, either within LDLR or in a neighboring gene. The minor allele of rs1433099 is common (MAF range: 26.5% to 34.8%) across 5 ethnic groups. The PAR for MI of the 9 INTERHEART risk factors is 89.6%, and the PAR of the 9 risk factors plus rs1433099 is 91.2%, an increase of 1.6%.
Of the 9 modifiable risk factors originally defined in the INTERHEART case control study, the Apo B/A1 ratio accounted for most MI. Our current results are consistent with this, as the largest number of significant genetic associations were with the Apo B/A1 ratio. Thus, many of the genetic variants with the greatest influence on MI will most likely be mediated through the risk factors that have been shown to have the strongest effect on MI.
Our analysis using detailed information collected on intermediate phenotypes to explore putative genes has demonstrated the impact of common variants on lipid levels and acute MI in a multiethnic sample. The large size of the multiethnic INTERHEART case-control study enabled us to clearly demonstrate that multiple SNPs that effect Apo B also contribute to MI, whereas CETP SNPs that have a strong effect on Apo A1 are not associated with MI risk. Specifically, we demonstrated that the Apo E 2 isoform, and 2 common LDLR variants (rs1433099 and rs6511720) influence MI risk and 1 LDLR SNP (rs1433099) contributes to MI risk independent of all risk factors, including lipid levels.
The authors acknowledge the contributions of all INTERHEART investigators who worked tirelessly to maximize recruitment and ensure high quality data. The authors thank the study participants. The authors also thank Swneke D. Bailey, Ron Do, and Lise Coderre for helpful discussions.
Drs Anand and Engert coordinated the data analysis and wrote the manuscript. Dr Xie performed the primary statistical analysis. Dr Montpetit coordinated all aspects of genotyping and quality control. Dr Paré provided advice on quality control and analytic strategy. Dr Hudson obtained funding for the study, and oversaw the genotyping and quality control. Ms Rangarajan coordinated the INTERHEART study. Dr McQueen supervised the Clinical Research Laboratory. Drs Cordell and Keavney provided advice on the statistical analysis. Dr Yusuf initiated and supervised the conduct of the main INTERHEART study, and obtained funding for the study. All authors critically reviewed the manuscript.
Dr Sonia Anand, Dr Changchun Xie, and Dr Jamie Engert had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Sources of Funding
This work was supported by Genome Canada, Genome Quebec; and the Population Health Research Institute, Hamilton, Canada.
World Health Organization 2006. Available at: http://www.who.int/cardiovascular_diseases/priorities/en/. Data last accessed: August 1, 2008.
Murray CJL, Lopez AD. Quantifying the burden of disease and injury attributable to ten major risk factors. In: Murray CJL, Lopez AD, eds. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability From Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Cambridge, MA: Harvard University Press; 1996.
Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lisheng L; INTERHEART Study Investigators. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the interheart study): case-control study. Lancet. 2004; 364: 937–952.
McQueen MJ, Hawken S, Wang X, Ounpuu S, Sniderman A, Probstfield J, Steyn K, Sanderson JE, Hasani M, Volkova E, Kazmi K, Yusuf S; INTERHEART Study Investigators. Lipids, lipoproteins, and apolipoproteins as risk markers of myocardial infarction in 52 countries (the INTERHEART study): a case-control study. Lancet. 2008; 372: 224–233.
Willer CJ, Sanna S, Jackson AU, Scuteri A, Bonnycastle LL, Clarke R, Heath SC, Timpson NJ, Najjar SS, Stringham HM, Strait J, Duren WL, Maschio A, Busonero F, Mulas A, Albai G, Swift AJ, Morken MA, Narisu N, Bennett D, Parish S, Shen H, Galan P, Meneton P, Hercberg S, Zelenika D, Chen WM, Li Y, Scott LJ, Scheet PA, Sundvall J, Watanabe RM, Nagaraja R, Ebrahim S, Lawlor DA, Ben-Shlomo Y, Davey-Smith G, Shuldiner AR, Collins R, Bergman RN, Uda M, Tuomilehto J, Cao A, Collins FS, Lakatta E, Lathrop GM, Boehnke M, Schlessinger D, Mohlke KL, Abecasis GR. Newly identified loci that influence lipid concentrations and risk of coronary artery disease. Nat Genet. 2008; 40: 161–169.
McPherson R, Pertsemlidis A, Kavaslar N, Stewart A, Roberts R, Cox DR, Hinds DA, Pennacchio LA, Tybjaerg-Hansen A, Folsom AR, Boerwinkle E, Hobbs HH, Cohen JC. A common allele on chromosome 9 associated with coronary heart disease. Science. 2007; 316: 1488–1491.
Kathiresan S, Melander O, Guiducci C, Surti A, Burtt NP, Rieder MJ, Cooper GM, Roos C, Voight BF, Havulinna AS, Wahlstrand B, Hedner T, Corella D, Tai ES, Ordovas JM, Berglund G, Vartiainen E, Jousilahti P, Hedblad B, Taskinen MR, Newton-Cheh C, Salomaa V, Peltonen L, Groop L, Altshuler DM, Orho-Melander M. Six new loci associated with blood low-density lipoprotein cholesterol, high-density lipoprotein cholesterol or triglycerides in humans. Nat Genet. 2008; 40: 189–197.
Yusuf S, Hawken S, Ounpuu S, Bautista L, Franzosi MG, Commerford P, Lang CC, Rumboldt Z, Onen CL, Lisheng L, Tanomsup S, Wangai P Jr., Razak F, Sharma AM, Anand SS; INTERHEART Study Investigators. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet. 2005; 366: 1640–1649.
Paré G, Serre D, Brisson D, Anand SS, Montpetit A, Tremblay G, Engert JC, Hudson TJ, Gaudet D. Genetic analysis of 103 candidate genes for coronary artery disease and associated phenotypes in a founder population reveals a new association between endothelin-1 and high-density lipoprotein cholesterol. Am J Hum Genet. 2007; 80: 673–682.
Fan JB, Oliphant A, Shen R, Kermani BG, Garcia F, Gunderson KL, Hansen M, Steemers F, Butler SL, Deloukas P, Galver L, Hunt S, McBride C, Bibikova M, Rubano T, Chen J, Wickham E, Doucet D, Chang W, Campbell D, Zhang B, Kruglyak S, Bentley D, Haas J, Rigault P, Zhou L, Stuelpnagel J, Chee MS. Highly parallel SNP genotyping. Cold Spring Harb Symp Quant Biol. 2003; 68: 69–78.
Ehrich M, Bocker S, van den Boom D. Multiplexed discovery of sequence polymorphisms using base-specific cleavage and MALDI-TOF MS. Nucleic Acids Res. 2005; 33: e38.
Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000; 155: 945–959.
Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003; 327: 557–560.
Samani NJ, Erdmann J, Hall AS, Hengstenberg C, Mangino M, Mayer B, Dixon RJ, Meitinger T, Braund P, Wichmann HE, Barrett JH, König IR, Stevens SE, Szymczak S, Tregouet DA, Iles MM, Pahlke F, Pollard H, Lieb W, Cambien F, Fischer M, Ouwehand W, Blankenberg S, Balmforth AJ, Baessler A, Ball SG, Strom TM, Braenne I, Gieger C, Deloukas P, Tobin MD, Ziegler A, Thompson JR, Schunkert H; WTCCC and the Cardiogenics Consortium. Genome wide association analysis of coronary artery disease. N Engl J Med. 2007; 357: 443–453.
Keavney B, Palmer A, Parish S, Clark S, Youngman L, Danesh J, McKenzie C, Delépine M, Lathrop M, Peto R, Collins R; International Studies of Infarct Survival (ISIS) Collaborators. Lipid-related genes and myocardial infarction in 4685 cases and 3460 controls: discrepancies between genotype, blood lipid concentrations, and coronary disease risk. Int J Epidemiol. 2004; 33: 1002–1013.
Gomaraschi M, Baldassarre D, Amato M, Eligini S, Conca P, Sirtori CR, Franceschini G, Calabresi L. Normal vascular function despite low levels of high-density lipoprotein cholesterol in carriers of the apolipoprotein A-I (Milano) mutant. Circulation. 2007; 116: 2165–2172.
Osier MV, Pakstis AJ, Soodyall H, Comas D, Goldman D, Odunsi A, Okonofua F, Parnas J, Schulz LO, Bertranpetit J, Bonne-Tamir B, Lu RB, Kidd JR, Kidd KK. A global perspective on genetic variation at the ADH genes reveals unusual patterns of linkage disequilibrium and diversity. Am J Hum Genet. 2002; 71: 84–99.
Hines LM, Stampfer MJ, Ma J, Gaziano JM, Ridker PM, Hankinson SE, Sacks F, Rimm EB, Hunter DJ. Genetic variation in alcohol dehydrogenase and the beneficial effect of moderate alcohol consumption on myocardial infarction. N Engl J Med. 2001; 22: 344:549–555.
Ebrahim S, Lawlor DA, Shlomo YB, Timpson N, Harbord R, Christensen M, Baban J, Kiessling M, Day I, Gaunt T, Davey Smith G. Alcohol dehydrogenase type 1C (ADH1C) variants, alcohol consumption traits, HDL-cholesterol and risk of coronary heart disease in women and men: British Women’s Heart and Health Study and Caerphilly cohorts. Atherosclerosis. 2008; 196: 871–878.
Brown MS, Goldstein JL. Expression of the familial hypercholesterolemia gene in heterozygotes: mechanism for a dominant disorder in man. Science. 1974; 185: 61–63.
Muallem H, North KE, Kakoki M, Wojczynski MK, Li X, Grove M, Boerwinkle E, Wilhelmsen KC, Heiss G, Maeda N. Quantitative effects of common genetic variations in the 3′UTR of the human LDL-receptor gene and their associations with plasma lipid levels in the Atherosclerosis Risk in Communities study. Hum Genet. 2007; 121: 421–431.
Myocardial infarction (MI) is the leading cause of death world-wide. The INTERHEART case-control study showed that nine risk factors (dyslipidemia, diabetes, hypertension, abdominal obesity, tobacco exposure, physical inactivity, psychosocial stressors, low fruit and vegetable intake, and no alcohol consumption) contribute to the risk for MI globally. The ratio of apolipoprotein B/A1 was the strongest MI risk factor and accounted for 54% of the total population attributable risk (PAR) of MI. The current paper identified 11 single nucleotide polymorphisms (SNPs) that were strongly associated with the ApoB/A1 ratio; eight SNPs from three genes are associated with ApoB, and three CETP SNPs were associated with ApoA1. Three of these SNPs are significantly associated with MI and one SNP in the LDLR gene is independently associated with MI after adjustment for all nine INTERHEART risk factors. While the SNPs that were associated with ApoB were also associated with MI, this was not the case for the CETP SNPs which were very significantly associated with ApoA1. This suggests that not all variation of ApoA1 or HDL will result in an alteration of MI risk. In addition, we demonstrate unequivocally that both the ε2 and the ε4 isoforms of ApoE effect MI risk. Our research suggests that in the future it may be possible to utilize an individual’s genetic profile to aid in their clinical management of dyslipidemia and MI.
The online Data Supplement is available at http://circgenetics.ahajournals.org/cgi/content/full/CIRCGENETICS.108.813709/DC1.