Author + information
- Received October 2, 2007
- Revision received February 8, 2008
- Accepted March 25, 2008
- Published online July 29, 2008.
- Maria C.A. Stene, MSc, PhD⁎,
- Ruth Frikke-Schmidt, MD, PhD⁎,
- Børge G. Nordestgaard, MD, DMSc‡,§,
- Peer Grande, MD, DMSc†,
- Peter Schnohr, MD§ and
- Anne Tybjærg-Hansen, MD, DMSc⁎,§,⁎ ()
- ↵⁎Reprint request and correspondence:
Dr. Anne Tybjærg-Hansen, Chief Physician, Associate Professor, Department of Clinical Biochemistry KB3011, Section for Molecular Genetics, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 9, DK-2100 Copenhagen, Denmark.
Objectives This study was designed to test the hypotheses that single nucleotide polymorphisms (SNPs), in zinc finger protein 202 (ZNF202), predict severe atherosclerosis and ischemic heart disease (IHD).
Background ZNF202 is a transcriptional repressor controlling promoter elements in genes involved in vascular maintenance and lipid metabolism.
Methods We first determined genotype association for 9 ZNF202 SNPs with severe atherosclerosis (ankle brachial index >0.7 vs. ≤0.7) in a cross-sectional study of 5,355 individuals from the Danish general population. We then determined genotype association with IHD in 10,431 individuals from the Danish general population, the CCHS (Copenhagen City Heart Study), including 1,511 incident IHD events during 28 years of follow-up. Results were verified in 2 independent case-control studies including, respectively, 942 and 1,549 cases with IHD and 8,998 controls. Finally, we determined whether g.−660A>G altered transcriptional activity of the ZNF202 promoter in vitro.
Results Cross-sectionally, ZNF202 g.−660 GG versus AA homozygosity predicted an odds ratio for severe atherosclerosis of 2.01 (95% confidence interval [CI]: 1.34 to 3.01). Prospectively, GG versus AA homozygosity predicted a hazard ratio for IHD of 1.21 (95% CI: 1.02 to 1.43). In the 2 case-control studies, the equivalent odds ratios for IHD were 1.29 (95% CI: 1.02 to 1.62) and 1.60 (95% CI: 1.34 to 1.92), confirming the results from the prospective study. Only 2 other SNPs, which were highly correlated with g.−660A>G, also predicted risk of severe atherosclerosis and IHD. Finally, ZNF202 g.−660G versus g.−660A was associated with a 60% reduction in transcriptional activity in vitro, whereas none of the 2 correlated SNPs were predicted to be functional.
Conclusions Homozygosity for a common functional promoter variant in ZNF202 predicts severe atherosclerosis and an increased risk of IHD.
Atherosclerosis is 1 of the leading causes of death and disability in the Western world, accounting for up to 50% of all deaths (1). Atherosclerosis is both a disease entity and the primary cause of ischemic heart disease (IHD). Epidemiological studies have identified numerous risk factors for atherosclerosis, including family history, gender, obesity, smoking, diabetes, hypertension, systemic inflammation, and dyslipidemia (1,2). Twin and family studies suggest that the heritability of atherosclerosis is about 50% (1,3). Several candidate genes have been identified; however, these genes only explain a fraction of the genetic component of atherosclerosis (1,4). For this reason, the search for yet other genes potentially involved in the development of atherosclerosis, and thus IHD, is important.
Linkage analysis of Utah pedigrees with heritable hypoalphalipoproteinemia and a family history of early coronary heart disease identified a susceptibility locus on chromosome 11q23 where zinc finger protein 202 (ZNF202) is localized (5,6). ZNF202 binds a promoter element found in vascular endothelial growth factor (VEGF) (7) involved in blood vessel maintenance (8) and also controls promoter elements found in several genes involved in lipid metabolism (7,9,10). This suggests that ZNF202 might be important for the development of atherosclerosis and IHD.
In a previous study (11,12), we resequenced the promoter and coding region of ZNF202 in 190 individuals and identified 17 genetic variants, 10 of which were classified as single nucleotide polymorphisms (SNPs) (minor allele frequency >1%; the 10th SNP, p.V244V, was not typed). In the present study, we tested the following hypotheses: 1) SNPs in ZNF202 predict risk of severe atherosclerosis as determined by ankle brachial index (ABI) in a large cross-sectional study of the Danish general population; 2) SNPs in ZNF202 predict risk of IHD in a large prospective study of the Danish general population, and this finding was replicated in 2 independent case-control studies; and 3) ZNF202 g.−660A>G is a functional promoter SNP in vitro.
General Population Sample
The CCHS (Copenhagen City Heart Study) is a prospective cardiovascular study of the Danish general population initiated from 1976 to 1978 with 3 follow-up examination periods—1981 to 1983, 1991 to 1994, and 2001 to 2003 (13,14). Individuals were selected based on the national Danish Civil Registration System to reflect the adult Danish general population ages 20 to 80+ years. Blood samples for DNA extraction were available on participants from the 1991 to 1994 and 2001 to 2003 examinations. Nine SNPs (g.−685G>A, g.−660A>G, g.−118G>T, g.+34G>A, c.IVS4–240A>T [NT_033899, nucleotide no. 11499], c.IVS4–223T>C [NT_033899, nucleotide no. 11516], p.A154V [c.461C>T], p.K259E [c.775A>G], c.*2T>G), previously identified by resequencing the ZNF202 promoter in 190 individuals of Danish ancestry (11), were genotyped in 8,942 individuals in the present and a previous study (11). The g.−660A>G SNP was genotyped in an additional 1,489 individuals (n = 10,431). Information on risk factors was available on all individuals genotyped.
Ankle brachial index (ABI), a blood pressure reduction in the legs that predicts severe atherosclerosis (15–18), was determined only in the approximately 6,000 participants in the 2001 to 2003 examination of the CCHS. Of these, 4,142 participants were genotyped for all 9 SNPs, and an additional 1,213 participants (n = 5,355) were genotyped for g.−660A>G. A standard brachial systolic and diastolic blood pressure was recorded on both arms, and systolic ankle blood pressure of the posterior tibial artery on both legs was obtained by Doppler (Huntleigh Mini Dopplex Doppler D900, Huntleigh, United Kingdom). The ABI was the lowest ankle systolic blood pressure divided by the highest brachial systolic blood pressure (15).
Information on diagnoses of IHD (World Health Organization International Classification of Diseases-8th edition, codes 410 to 414; -10th edition, codes I20 to I25) was collected and verified until the end of 2003 by reviewing all hospital admissions and diagnoses entered in the national Danish Patient Registry, all causes of death entered in the national Danish Causes of Death Registry, and medical records from hospitals and general practitioners. Ischemic heart disease was a previous myocardial infarction (MI) or characteristic symptoms of angina pectoris (non-MI IHD) based on location, character, and duration of pain, and the relation of pain to exercise (19). A diagnosis of MI required the presence of at least 2 of the following criteria: characteristic chest pain, elevated cardiac enzymes, and electrocardiographic changes indicative of an MI. The MI end points gathered from the national Danish Patient Registry and the national Causes of Death Registry were confirmed independently via medical records of general practitioners and hospitals. Angina pectoris was not confirmed independently, and therefore, some misclassification for angina and thus for the entire IHD group in the prospective study may be present.
Patients With IHD—Group 1
A second cohort comprised 942 patients from the greater Copenhagen area referred for coronary angiography to Copenhagen University Hospital, during the period from 1991 through 1994. These patients were evaluated by experienced cardiologists and had documented IHD based on characteristic symptoms of angina pectoris (19), plus at least 1 of the following: stenosis/atherosclerosis on coronary angiography, a previous MI, or a positive exercise electrocardiography test. The diagnosis of MI was established with the same criteria as in the general population sample.
Patients With IHD—Group 2
A third cohort comprised an additional 1,549 patients with IHD referred to coronary angiography at the same hospital during the period 2000 to 2004. The diagnosis of IHD was established by experienced cardiologists using the same criteria as previously mentioned.
Patients with IHD were genotyped for the 9 ZNF202 SNPs mentioned previously. Ankle-brachial index was not available in either of these patient groups.
Studies were approved by institutional review boards and Danish ethical committees: nos. KFV.100.2039/91 and KF01-144/01, Copenhagen and Frederiksberg committee, and KA93125 and KA99039, Copenhagen County committee, and conducted according to the Declaration of Helsinki. Informed consent was obtained from participants. More than 99% were white and of Danish descent.
Cross-Sectional Study of Risk of Severe Atherosclerosis
We included, respectively, 5,355 (g.−660A>G only) and 4,142 (remaining 8 SNPs) participants from the 2001 to 2003 examination of the CCHS in which ABI was available and determined whether ZNF202 genotypes were associated with an increased odds ratio for severe atherosclerosis by comparing genotype distribution in individuals from the general population with an ABI >0.7 versus ≤0.7. In addition, we compared g.−660A>G genotype distribution in individuals with extreme ABI cut-points (ABI >0.9 vs. ≤0.7, ABI >0.9 vs. ≤0.5).
Prospective Study of Risk of Ischemic Heart Disease
We included, respectively, 10,431 (g.−660A>G only) and 8,942 (8 SNPs) participants from the 1991 to 1994 and 2001 to 2003 examinations of the CCHS. All end points were recorded in the follow-up period 1976 through 2003. The follow-up time was 28 years (213.400 person-years). Individuals diagnosed with IHD before study entry were excluded. We observed, respectively, 1,511 (g.−660A>G only) and 1,433 (8 SNPs) incident IHD events. Because ABI was not available in the 1991 to 1994 examination of the CCHS, risk of severe atherosclerosis could not be determined in this study.
Case-Control Studies of Risk of Ischemic Heart Disease
Case-control study 1 included 942 cases with IHD and, respectively, 8,998 (g.−660A>G only) and 7,644 (8 SNPs) unmatched controls from the CCHS without IHD. Case-control study 2 included 1,549 cases with IHD and the same controls as previously mentioned. Because ABI was not available in either of these patient groups, risk of severe atherosclerosis could not be determined in these 2 studies.
The ABI PRISM 7900HT Sequence Detection System (Applied Biosystems Inc., Foster City, California) was used to genotype 9 ZNF202 SNPs identified by resequencing ZNF202 as previously described (a 10th SNP, p.V244V, was not typed) (Online Table 1) (11). All genotypes were in Hardy-Weinberg equilibrium (chi-square: p > 0.05).
Colorimetric and turbidimetric assays were used to measure plasma levels of total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, apolipoproteins B and AI (all Boehringer Mannheim GmbH, Mannheim, Germany).
The risk factors—diabetes mellitus, smoking, and hypertension—were dichotomized and defined as ever-diabetics (self-reported disease, use of insulin, use of oral hypoglycemic drugs and/or nonfasting plasma glucose >11.0 mmol/l), ever-smokers (ex-smoker or current smoker), ever-hypertensives (systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg and/or use of antihypertensive drugs). Body mass index was weight (kilograms) divided by height squared (square meters) and was tertilized.
ZNF202 G.−660G>A Expression Analysis
Reporter gene constructs for the ZNF202 promoter sequence (−710 to +124) were cloned into the HindIII and XhoI restriction sites of the pGL3-basic vector (Promega, Mannheim, Germany) and confirmed by sequence analysis. Mutagenesis was conducted using QuikChangeII Site-Directed Mutagenesis Kit (Stratagene, Amsterdam, the Netherlands) to incorporate the ZNF202 g.−660G SNP. Subsequently, the promoter insert with g.−660G was subcloned into a new pGL3-basic vector and confirmed by sequence analysis. HepG2 cells were maintained in culture in Dulbecco's Modified Eagle Medium (Gibco, Grand Island, New York) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin-streptomycin (Gibco) under 5% CO2 and at 37°C. The day before transfection, cells were seeded in 24-well plates at 1.8 × 105 cells per well. HepG2 cells were transiently transfected using FuGENE transfection reagent (Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer's protocol. To account for variable transfection efficiency, 0.5 μg of each ZNF202 promoter construct was cotransfected with 0.1 μg of RpL null vector (Promega). After 24 h of incubation, cells were harvested using Passive Lysis Buffer (Promega). The luciferase assay was performed using the dual-luciferase reporter assay system (Promega). Luciferase activity was measured using a TR717 microplate luminometer (Applied Biosystems, Bedford, Massachusetts). Three independent experiments were performed in triplicate with 2 different preparations of DNA. Results were expressed as relative luciferase activity.
We used the statistical software package Stata (Stata Corp., College Station, Texas). Two-sided probability values <0.05 were considered significant. In addition, p values were corrected for multiple comparisons for 9 SNPs using the Bonferroni adjustment method, which changed the required significance level from ≤0.05 to ≤0.006 (0.05/9 = 0.006). This correction is for the multiple SNP tests; however, it does not account for the multiple phenotypes and genetic models examined. Noncarriers, heterozygotes, and homozygotes were compared by Kruskal-Wallis analysis of variance for continuous variables and by Pearson chi-square test for categorical variables. In cross-sectional and case-control studies, logistic regression analysis adjusted for age in 10-year age groups was used to estimate odds ratios for severe atherosclerosis (ABI >0.7 vs. ≤0.7, ABI >0.9 vs. ≤0.7, ABI >0.9 vs. ≤0.5), IHD, MI, and angina pectoris as a function of genotypes. In the prospective study, with the use of left truncation (or delayed entry), Cox proportional hazards regression models with age as time scale estimated hazard ratios for IHD, MI, and angina pectoris. For all studies, cross-sectional, prospective, and case-control studies, for all end points, and for all 9 SNPs, the following a priori models of inheritance were assumed: an assumption-free model with the most common genotype as the reference group, as well as additive, recessive, and dominant models. Pairwise linkage disequilibrium between the 9 ZNF202 SNPs was estimated in participants in the CCHS using Haploview and expressed as D′ and r2 (20,21). Mean relative luciferase activity as a function of ZNF202 g.−660A>G genotype was compared by Mann-Whitney U test.
Genotype and minor allele frequencies for all 9 ZNF202 SNPs in the Danish general population are shown in Table 1. This section of this report focuses on the g.−660A>G SNP, while the full tables for all 9 SNPs for all models tested are shown as Online Tables 2 to 11.
For the present study, complete genotypes for all 9 SNPs (p.V244V not typed) were available in 8,942 individuals from the general population. The minor alleles of g.−660A>G, c.IVS4–223T>C, and p.A154V were in linkage disequilibrium (D′ = +0.99) and also highly correlated (r2= 0.98 to 0.99), indicating that they were on the same haplotype, and that 1 SNP could tag or serve as a proxy for the other SNPs (Fig. 1).
Risk of severe atherosclerosis
In the cross-sectional study, including 5,126 individuals with an ABI >0.7 and 229 individuals with an ABI ≤0.7, the age-adjusted odds ratio for having an ABI ≤0.7 predicting severe atherosclerosis was 2.01 (95% confidence interval [CI]: 1.34 to 3.01; p = 0.001) for g.−660 GG versus AA homozygotes (Fig. 2A,Table 2). After adjustment for age, gender, total cholesterol, HDL cholesterol, triglycerides, body mass index, smoking, diabetes mellitus, and hypertension, the equivalent odds ratio was 2.00 (95% CI: 1.33 to 3.02; p = 0.001). Similar results were observed using more extreme cut-off points for ABI (>0.9 vs. ≤0.7 and >0.9 vs. ≤0.5) (Figs. 2B and 2C), as well as for the tag SNPs c.IVS4–223T>C and p.A154V (Online Table 2). Results were similar assuming a recessive mode of inheritance (Online Table 2). The mean ABI values for g.−660AA, g.−660AG, and g.−660GG were 1.02, 1.01, and 1.00, respectively (ANOVA; p = 0.05).
Risk of IHD
In the prospective study, the age adjusted hazard ratio for IHD, was 1.21 (95% CI: 1.02 to 1.43; p = 0.03) for g.–660 GG versus AA homozygotes (Fig. 3, upper left panel; Table 2). After adjustment for age, gender, total cholesterol, HDL cholesterol, triglycerides, body mass index, smoking, diabetes mellitus, and hypertension, the equivalent hazard ratio was 1.19 (95% CI: 1.01 to 1.41; p = 0.04).
In the verification samples, case-control studies 1 and 2, the age-adjusted odds ratios for IHD were, respectively, 1.29 (95% CI: 1.02 to 1.62; p = 0.03) and 1.60 (95% CI: 1.34 to 1.92; p = 0.0000003), and the combined odds ratio was 1.48 (95% CI: 1.27 to 1.73; p = 0.0000006) (Fig. 3, upper middle and right panels; Table 2). After adjustment for age and gender, the equivalent odds ratios were, respectively, 1.26 (95% CI: 0.99 to 1.59; p = 0.06), 1.52 (95% CI: 1.26 to 1.83; p = 0.00001), and 1.41 (95% CI: 1.20 to 1.66, p = 0.00002). When IHD was subdivided into MI and angina pectoris, results for g.−660 GG versus AA were essentially similar for MI in all 3 studies, as well as for angina in the study with the largest number of cases, case-control study 2 (Fig. 3, middle and lower panels; Online Tables 6 to 9). With the exception of the 2 tag SNPs, c.IVS4–223 and p.A154V, none of the other 6 SNPs were consistently associated with an increased risk of IHD, MI, or angina pectoris (Online Tables 3 to 11). Results were similar assuming a recessive mode of inheritance (Online Tables 3 to 11). Finally, well-known risk factors for atherosclerosis and IHD, including lipids and lipoproteins, did not differ as a function of g.−660A>G genotypes (Table 3).
ZNF202 g.−660A>G promoter function
The relative luciferase activity driven by the ZNF202 promoter with the mutant G at position g.−660 was consistently reduced by more than 60% compared with the wild-type promoter with A at position g.−660 (Fig. 4) (p values from 0.01 to 0.004 by Mann-Whitney U test for 3 separate experiments each performed in triplicate and with 2 different DNA preparations). These results indicated that the G allele was associated with reduced transcriptional activity compared with the A allele.
The key findings in this study are: 1) ZNF202 g.−660 GG homozygotes (frequency: 9% in the general population) had an increased risk of severe atherosclerosis, as determined by a low ABI in a cross-sectional study of more than 5,000 individuals from the general population. 2) In a large prospective study of more than 10,000 individuals followed for 28 years with 1,511 incident IHD events, g.−660 GG homozygosity predicted an increased risk of IHD. The increased risk of IHD associated with GG homozygosity was confirmed in 2 large, independent case-control studies including a total of approximately 2,500 cases with IHD and 9,000 controls. 3) The g.−660A>G, G allele, was associated with decreased transcriptional activity in vitro.
ZNF202 is located in a susceptibility locus for familial hypoalphalipoproteinemia (5) and is a transcriptional repressor controlling promoter elements predominantly found in genes involved in vascular maintenance and lipid metabolism (7,9,10,22). For this reason, it is plausible that ZNF202 could be a candidate gene for atherosclerosis and risk of IHD. ZNF202 is known to bind a promoter element found in VEGF, suggesting that a possible mechanism behind our observations could be an effect of genetic variation in ZNF202 on the expression of VEGF. VEGF is a potent angiogenic factor that plays multiple roles in vascular development and maintenance (8). Several observations indicate that VEGF has proatherosclerotic properties. VEGF enhances atherosclerotic plaque progression (23,24) and induces plaque destabilization via neovascularization, which increases macrophage accumulation and progression of atherosclerosis (25). VEGF induces the expression of endothelial adhesion molecules and has chemoattractant properties, suggesting a role for VEGF in leukocyte recruitment (26,27). VEGF also stimulates tissue factor expression on endothelial cells and thereby the conversion of prothrombin to thrombin, enhancing a procoagulant state (28). An altered transcriptional activity of the ZNF202 promoter could result in altered VEGF expression, potentially explaining the observed atherosclerosis and IHD susceptibility.
Other ZNF202 target genes are involved in lipid metabolism (7,9,10,22). ZNF202 could also influence risk of atherosclerosis and IHD via regulation of 1 or more of these genes. ZNF202 g.−660A>G genotype was not associated with variation in plasma levels of total cholesterol, apolipoprotein B, low-density lipoprotein cholesterol, HDL cholesterol, apolipoprotein AI, and triglycerides (Table 3); however, this does not preclude an effect on target genes involved in lipid metabolism. In support of this, we have previously shown that a common mutation in the cholesterol transporter ABCA1, a well-known HDL gene and a target gene for ZNF202 (9,10), predicts an increased risk of IHD independent of lipid and lipoprotein levels in plasma (29).
ZNF202 g.−660A>G was associated with severe atherosclerosis with IHD and MI in all 3 studies and, in addition, with angina pectoris in case-control study 2, the study that included by far the largest number of cases with angina (>900 cases). These results might suggest an effect of g.−660A>G on the development of atherosclerotic plaque progression, as well as an effect on plaque stability and thrombosis. As noted above, the target gene VEGF may play a role in the development of all these phenotypes.
In a previous study (12), which included only SNPs in the coding region of ZNF202, a common nonsynonymous SNP, p.A154V, predicted an increased risk of IHD in participants of the CCHS. In the present study, we found that p.A154V was both in tight linkage disequilibrium and highly correlated with g.−660A>G and c.IVS4–223T>C, raising the question which of these was the functional SNP. Because p.A154V was not conserved between species (human, mouse, rat) and substituted similar amino acids, and because c.IVS4–223T>C was not predicted to affect splicing or gene regulation, this suggested that g.−660A>G could be the functional SNP. This was confirmed in in vitro expression studies consistently showing decreased transcriptional activity of the g.−660 G allele.
In the assumption-free model, the g.−660A>G genotype was entered as a categorical variable with AA as the reference group, and without an a priori assumption of mode of inheritance. This model was consistently significant only for the comparison of GG versus AA genotype. In agreement with this, and assuming different genetic models a priori (additive, recessive, dominant), a recessive model of inheritance was also the most consistently significant model throughout the studies.
Although our studies are very large, and both cross-sectional and prospective and are based on a sample of the general population, there are potential limitations. Misclassification of end points is always a concern; however, MI end points gathered from the national Danish Patient Registry and the national Danish Causes of Death Registry were confirmed independently via medical records of general practitioners and hospitals. Angina pectoris was not confirmed independently, and therefore, some misclassification for angina and thus for the entire IHD group in the prospective study may be present. Despite this, genotypes predicted IHD suggesting that the true association might be stronger than that observed. In support of this possibility, the odds ratios for IHD in the 2 case-control studies were both larger than the hazard ratios for IHD observed in the prospective study: IHD cases in the 2 case-control studies were confirmed by stenosis/atherosclerosis on coronary angiography, a previous MI or a positive exercise electrocardiography test. Misclassification of genotypes is very unlikely, because genotype distribution in the general population was in Hardy-Weinberg equilibrium, because genotyping was performed using reliable TaqMan chemistry, and because all data transfer from genotyping equipment to the database was performed electronically, excluding human error in database entry.
Another concern in this type of study with multiple comparisons is type I error. One of the currently most widely accepted ways of dealing with type I error is to replicate one's results in different samples. A second way of addressing type I error is to correct for multiple comparisons using 1 of several statistical methods. In this study, we used both approaches. We observed similar results for 3 related end points—severe atherosclerosis, IHD, and MI—in 2 large studies and confirmed these results for IHD and MI in 2 additional, large independent studies, suggesting that these results were not due to type I error. Even after correction for multiple comparisons using the Bonferroni adjustment method, risk of severe atherosclerosis, IHD, MI, and angina pectoris remained significant in 2 studies.
We show that homozygosity for a common functional promoter variant in ZNF202, g.−660G, predicts severe atherosclerosis and an increased risk of IHD. This suggests that ZNF202 is a candidate gene for atherosclerosis and IHD.
The authors thank Karin Møller Hansen, Mette Refstrup, and Ragnhild Aanestad for excellent technical assistance. We are indebted to the staff and participants of the Copenhagen City Heart Study, and to the patients with IHD for their important contributions.
For supplementary tables, please see the online version of this article.
This study was supported by grants from the Danish Medical Research Council, the Danish Heart Foundation, the Research Fund at Rigshospitalet, Copenhagen University Hospital, Ingeborg and Leo Dannin's grant, Helge Hansen's and wife's grant, all Copenhagen, Denmark; a Specific Targeted Research Project grant from the European Union, Sixth Framework Programme Priority ([FP-2005-LIFESCIHEALTH-6], contract #037631).
- Abbreviations and Acronyms
- ankle brachial index
- high-density lipoprotein
- ischemic heart disease
- myocardial infarction
- single nucleotide polymorphism
- Received October 2, 2007.
- Revision received February 8, 2008.
- Accepted March 25, 2008.
- American College of Cardiology Foundation
- Wagner S.,
- Hess M.A.,
- Ormonde-Hanson P.,
- et al.
- Hoeben A.,
- Landuyt B.,
- Highley M.S.,
- Wildiers H.,
- Van Oosterom A.T.,
- De Bruijn E.A.
- Porsch-Ozcurumez M.,
- Langmann T.,
- Heimerl S.,
- et al.
- Langmann T.,
- Schumacher C.,
- Morham S.G.,
- et al.
- Stene M.C.,
- Frikke-Schmidt R.,
- Nordestgaard B.G.,
- Tybjærg-Hansen A.
- Schnohr P.,
- Jensen G.,
- Lange P.,
- Scharling H.,
- Appleyard A.
- Appleyard M.,
- Tybjærg-Hansen A.,
- Jensen G.,
- Schnohr P.,
- Nyboe J.,
- Copenhagen City Heart Study Group
- Eldrup N.,
- Sillesen H.,
- Prescott E.,
- Nordestgaard B.G.
- Newman A.B.,
- Shemanski L.,
- Manolio T.A.,
- et al.,
- Cardiovascular Health Study Group
- Julian D.G.,
- Bertrand M.E.,
- Hjalmarson A.,
- Fox K.,
- Simoons M.L.,
- Task Force of the European Society of Cardiology
- Barrett J.C.,
- Fry B.,
- Maller J.,
- Daly M.J.
- Celletti F.L.,
- Hilfiker P.R.,
- Ghafouri P.,
- Dake M.D.
- Moulton K.S.,
- Vakili K.,
- Zurakowski D.,
- et al.
- Yamada M.,
- Kim S.,
- Egashira K.,
- et al.
- Kim I.,
- Moon S.O.,
- Kim S.H.,
- Kim H.J.,
- Koh Y.S.,
- Koh G.Y.
- Zucker S.,
- Mirza H.,
- Conner C.E.,
- et al.
- Frikke-Schmidt R.,
- Nordestgaard B.G.,
- Schnohr P.,
- Steffensen R.,
- Tybjærg-Hansen A.