You are here

Article Text

Article menu
Cardiac risk factors and prevention
Original article
Traditional risk factors and a Genetic Risk Score are associated with age of first acute coronary syndrome
  1. Christopher Labos1,
  2. Rui Hao Leo Wang2,
  3. Louise Pilote2,
  4. Peter Bogaty3,
  5. James M Brophy1,2,
  6. James C Engert4,
  7. George Thanassoulis2
  1. 1Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Quebec, Canada
  2. 2Department of Medicine, McGill University, and the Research Institute of the McGill University Health Center Montreal, Quebec, Canada
  3. 3Institut Universitaire de Cardiologie et de Pneumologie, Laval University, Quebec City, Quebec, Canada
  4. 4Department of Human Genetics, McGill University, Montreal, Quebec, Canada
  1. Correspondence to Dr George Thanassoulis, Preventive and Genomic Cardiology, McGill University Health Center, 687 Pine Ave W. H4.55, Montreal, Quebec, Canada H3A 1A1; george.thanassoulis{at}mcgill.ca

Abstract

Objective To examine the association between traditional risk factors (TRF) and a Genetic Risk Score (GRS) with age of first acute coronary syndrome (ACS). Early onset ACS may occur due to a high burden of TRFs or to genetic factors that accelerate atherosclerosis. Whether recently discovered genetic variants for ACS are more prevalent at earlier age of first ACS remains unknown.

Methods To construct a multilocus GRS, participants were genotyped for 30 single nucleotide polymorphisms (SNP) identified from prior genome-wide association studies. Linear regression models were fit to estimate the association between TRFs and GRS with age of first ACS.

Results We included 460 participants with a first ACS enrolled in the Recurrence and Inflammation in the Acute Coronary Syndromes (RISCA) cohort. Several TRFs were associated (all p<0.05) with earlier age of first ACS: male sex (6.9 years earlier (95% CI 4.1 to 9.7)), current cigarette smoking (8.1 years (95% CI 6.1 to 10.0)), overweight (Body Mass Index, BMI >25) and obesity (BMI>30) (5.2 years (95% CI 2.6 to 7.9)). In women, hormone replacement therapy was also associated with earlier age of first ACS (4.8 years earlier (95% CI 0.3 to 8.4)). After multivariable adjustment for TRFs, a 1 SD increment in the GRS was associated with a 1.0 (95% CI 0.1 to 2.0) year earlier age of first ACS.

Conclusions Among individuals with a first ACS, a GRS composed of 30 SNPs is associated with younger age of presentation. Although genetic predisposition modestly contributes to earlier ACS, a heavy burden of TRF is associated with markedly earlier ACS.

http://dx.doi.org/10.1136/heartjnl-2013-305416

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Premature myocardial infarction (MI), defined as an MI occurring prior to age 55 years, affects 2.2% of men and 1.0% of women annually.1 Given the typical trajectory of traditional risk factors for MI, an early onset of MI is frequently attributed to the possible presence of unique genetic factors that may accelerate the atherosclerotic process or predispose to a final step in the causal pathway of MI (ie, thrombosis or plaque rupture). However, early MI may also occur due to a high burden of traditional risk factors. Traditional risk factors, such as smoking, hypertension, dyslipidemia, and diabetes have well-established roles in the development of coronary artery disease,2–4 and several such factors have been associated with earlier age of first MI.5–8

    Family history of MI is a marker of genetic risk9 that can be easily ascertained and is associated with earlier MI,7 but remains a crude marker of genetic risk. High-throughput genotyping of large samples has permitted the discovery of many single nucleotide polymorphisms (SNP) that are robustly associated with MI.10–18 However, little data exists regarding the association between these SNPs and age of first MI.19–21 Additionally, the effect of a Genetic Risk Score (GRS) on age of first MI has not been investigated. Accordingly, we examined the impact of traditional risk factors as well as a multilocus GRS composed of SNPs strongly associated with MI from large-scale genome-wide association studies on the age of a first acute coronary syndrome (ACS).

    Methods

    Study sample

    The study participants were from the previously published RISCA (Recurrence and Inflammation in the Acute Coronary Syndromes) study.22 Briefly, 1210 consecutive patients were recruited from four tertiary and four Canadian community hospitals (seven in Quebec and one in New Brunswick) from 2001 until 2002. To be eligible, patients had to have an urgent admission with a diagnosis of either acute MI or unstable angina. All demographic and clinical data including traditional risk factors, results of all diagnostic and therapeutic procedures (including biochemical tests) and all medications taken prior to admission and prescribed at discharge were recorded, verified for consistency and systematically assessed by on-site visits.

    For the purposes of our analysis, only patients presenting with a first ACS comprising ST elevation and non-ST elevation MIs, as well as unstable angina were eligible. Therefore, of the 1210 individuals enrolled in RISCA, we excluded all patients with a prior history of MI (n=341). To ensure we included only individuals with their first cardiovascular event, we also excluded any patients with a history of coronary artery bypass grafting (n=70), percutaneous coronary intervention (n=82), stable and unstable angina, (n=121), stroke (n=18), heart failure (n=8), or prior cardiovascular admission (n=49). Thus, this study cohort consists of patients in whom ACS was the inaugural cardiac event (n=521). An additional 61 individuals were excluded from the analysis for missing data (n=59 for individuals who did not consent to genetic testing and n=2 for missing covariates).

    Outcome and covariate definitions

    The primary outcome was age at first ACS, as recorded at the time of enrolment in the RISCA study. A detailed description of the inclusion criteria for the RISCA study are included in the online appendix (see online supplement 1). Patients were defined as having diabetes if they had a history of diabetes (not including glucose intolerance) or were prescribed oral hypoglycaemic agents or insulin. Similarly, patients were defined as having hypertension or hypercholesterolaemia if they had a history of hypertension or hypercholesterolaemia or were prescribed antihypertensive therapy or lipid lowering therapy, respectively. Current smokers were defined as patients who continued to smoke (>1 cigarette per day) at the time of enrolment, or who had quit within the past 30 days. Body Mass Index (BMI) was categorised into three groups: normal if BMI ≤25, overweight if BMI >25 but ≤30, and obese if BMI >30. History of acetylsalicylic acid (ASA) use was defined as ASA therapy up to 325 mg daily and hormone replacement therapy (HRT) as any hormonal therapy prescribed to a menopausal woman (excluding any treatment for a malignancy) at the time of ACS admission.

    Genotyping and calculation of the GRS

    DNA extraction and genotyping was performed using standard techniques. Details are available in the online appendix (see online supplement 1). The GRS was determined a priori and constructed using genotypes from 30 uncorrelated SNPs (R2<0.3 between included SNPs) robustly associated and replicated in published genome-wide association studies (GWAS) of MI or coronary artery disease.9–18 As performed in prior work,23 a score for each individual was calculated as the unweighted sum of each risk allele across all 30 SNPs (ie, score of 2 for those homozygous for the risk allele, a score of 1 for heterozygotes, and a score of zero for the absence of the risk allele). Missing genotypes (<0.35% of all genotypes) were assumed to be missing at random (ie, non-informative missingness) and were imputed as two times the risk allele frequency, using the risk allele frequencies from the entire dataset. Thus, every individual could have a GRS ranging from 0 to 60; the actual range observed was 18 to 40.

    Statistical analysis

    Continuous variables were reported as means with SDs. Categorical variables were reported as counts with proportions. The association between age of first ACS and all variables was assessed using multivariable linear regression. The GRS was considered as a continuous variable centred about the mean. Linear regression models were fitted with age at first ACS as the dependent variable, and GRS as the independent variable. The following covariates were included in the adjusted model: sex, hypertension, diabetes, hypercholesterolaemia, current smoking, BMI (categorised as normal, overweight and obese), ASA use and HRT use. The output of the regression model was reported as β-coefficients representing the difference in the age of first ACS in years for each unit change of GRS or presence/absence of the covariate. To describe the risk factor burden and GRS scores across categories of age of first ACS, we divided the entire cohort into quintiles and calculated the mean number of risk factors and the GRS in each quintile. Means across quintiles were compared using one-way analysis of variance.

    In secondary analyses, we examined each SNP comprising the GRS individually with age of ACS. We also conducted a sensitivity analysis of our main analysis including only patients with MI (ie, excluding unstable angina). A p value of 0.05 was considered significant for all analyses except in exploratory analyses where each SNP was evaluated separately. For these per SNP associations, we considered a Bonferroni corrected p value of 0.002 (0.05/30 SNPs) to declare significance. All statistical testing was performed using STATA V.12 (StataCorp, College Station, Texas, USA).

    Results

    A total of 460 individuals (mean age 59±12 years, 22.4% female) were included in the final analysis. MI was the admitting diagnosis for 360 (78.3%) individuals, of which there were 238 STEMIs. The remainder were diagnosed as having unstable angina. In the study sample, 32.6% had hypertension, 12.4% had diabetes, and 37.8% had hypercholesterolaemia. Mean BMI was 27.0±4.4 and 40.2% of the population were current smokers. The mean GRS was 31.5±3.4. Full details of the baseline characteristics are presented in table 1.

    View this table:
    Table 1

    Baseline characteristics

    Associations between traditional risk factors and age of first ACS

    In fully adjusted models (all β coefficients expressed in years), male sex, current smoking status, HRT use, and being either overweight (BMI>25) or obese (BMI>30) were all statistically significantly associated (p<0.05) with younger age at first ACS (table 2). ASA use and hypertension were both associated with older age of first ACS. (table 2) The number of traditional risk factors present (male sex, hypertension, diabetes, hypercholesterolaemia, smoking, and obesity) was also strongly associated with the age of first ACS with a greater risk factor burden being associated with a younger age of first ACS, (β per additional risk factor=−2.2 (95% CI −3.2 to −1.3) p<0.001). A sensitivity analysis excluding unstable angina cases (ie, limited only to first MI cases) demonstrated no differences from the primary analysis.

    View this table:
    Table 2

    Linear associations between GRS and risk factors with age of first acute coronary syndrome

    Association between GRS and age of first ACS

    In unadjusted analyses, the GRS was associated with a younger age at first ACS, (b-coefficient −0.4 years per unit change in GRS (95% CI −0.7 years to −0.1 years)). In the fully adjusted model, the GRS remained statistically significantly associated with a younger age at first ACS, (β=−0.3 years (95% CI −0.6 years to −0.01 years)), as did the comparison between the highest versus the lowest quintile of GRS, β=−3.0 years (95% CI −5.9 years to −0.03 years). A 1 SD change (equivalent to 3.4 points) in the GRS was associated with an age at first ACS that was 1.0 years (95% CI 0.1 to 2.0) earlier (figure 1). Sensitivity analysis excluding unstable angina did not materially change these findings.

    Figure 1
    Figure 1

    Association of risk factors and Genetic Risk Score (GRS) with age at first ACS. *Difference in mean age for presence of risk factor as compared with absence. For BMI, the reference category was BMI <25. The GRS is presented per SD increase. Negative values indicate earlier mean age at first ACS. All p values <0.05. ASA, acetylsalicylic acid; HTN, hypertension; BMI, Body Mass Index; HRT, hormone replacement therapy; ACS, acute coronary syndrome.

    In exploratory analyses, we examined the association between individual SNPs that comprised the GRS and age of ACS; two SNPs were nominally statistically significant in multivariable analysis (at p<0.05): rs4977574 from chromosome 9p21 and rs9818870 from chromosome 3q22.3, but did not reach prespecified thresholds for significance due to multiple testing (see online supplementary table). The presence of each risk allele of 9p21 was associated with an age of first ACS that was 1.4 (95% CI 0.1 to 2.7) years earlier and each risk allele of 3q22.3 was associated with an age of first ACS that was 1.9 (0.1 to 3.7) years earlier.

    Risk factors and GRS by age of ACS

    To characterise the risk factor burden and GRS of earlier ACS cases as compared to later ACS, we examined the mean number of risk factors and the mean GRS across quintiles of age at first presentation. We observed a linear trend toward a higher number of risk factors (p<0.001) and a higher GRS with younger age of first ACS (p=0.011; figure 2). As compared with the oldest ACS cases (5th quintile; mean age=78.3), the earliest ACS cases (1st quintile; mean age=45.8 years) had a statistically higher mean number of risk factors (2.4 risk factors vs 1.9 risk factors; p=0.01) and GRS (31.9 vs 30.4 points; p=0.038).

    Figure 2
    Figure 2

    Mean number of risk factors and GRS per quintile of age of first ACS. GRS, Genetic Risk Score.

    Discussion

    In our study of 460 individuals with a first ACS, a higher GRS was statistically significantly associated with a younger age at first ACS. We found that for each SD increment in GRS (3.4 units), age of first ACS was 1 year earlier. We also show that in the highest quintile of GRS, the mean age of first ACS was nearly 3 years earlier as compared to the age in the lowest quintile of GRS. We demonstrate that two individual SNPs in the GRS (at the 9p21 and 3q22 loci) were associated with statistically significantly earlier ACS by 1.4 and 1.9 years per allele, respectively (see online supplementary table S1). Although these SNPs did not meet stringent criteria for significance due to multiple testing, our results corroborate previously reported associations between 9p21with a younger age of coronary artery disease onset, and provide independent replication of these findings.19–21 Last, we also confirm the association between several traditional risk factors and age of first ACS. Most notably, we found that current smoking and obesity were associated with markedly earlier first ACS (8 years and 5 years, respectively). Our results demonstrate that common genetic predisposition, as measured by a GRS, is only modestly associated with earlier presentation, whereas a heavy burden of traditional risk factors appears to be more strongly linked to earlier ACS. These results reinforce the need to promote risk factor reduction, especially cigarette smoking cessation and weight loss in young adults to prevent cardiovascular events at an earlier age.

    There is very limited data regarding the role of genetic factors on age of ACS despite the strong expectation of a genetic effect in earlier ACS. Madala et al7 demonstrated that family history of cardiovascular disease was associated with an earlier MI, but did not examine genetic data. Patel et al24, showed an association between a GRS and presence of coranary artery disease (CAD) in subjects under 70 years of age, but did not examine age of MI as an endpoint. Thus, to our knowledge, our study is the first to specifically evaluate the impact of GRS as a correlate of age at first ACS. Coronary heart disease in young individuals is usually assumed to have a genetic cause,25 and while a family history of heart disease has been associated with premature MI, 26 ,27 many traditional risk factors also play a role. In several prior studies, smoking was a powerful predictor of age at first MI, with smokers having their first MI 9.2, 9.3 and 9.7 years earlier, respectively.5–7 This is largely in agreement with our results which showed smoking was associated with an age of first ACS that is 8.1 years earlier as compared to non-smokers. BMI class has also been strongly associated with younger age at first MI. As compared to being of normal weight, being overweight (BMI between 25 and 30) was associated with an ACS 2.6 years earlier in our study, which agrees with the data from Bahler et al that also showed a 2.6-year earlier MI, and the data from Madala et al that showed a roughly 3-year earlier MI. Likewise, obesity (BMI greater than 30), was associated with a 5.2-year earlier ACS in our data and roughly 6.5 years earlier in the study by Madala et al Furthermore, they were able to show a clear linear relationship between age of first MI and increasing BMI levels into the extreme range above a BMI of 40.

    Although the effect of the GRS on earlier ACS was modest (1 year earlier per SD of a GRS), this is within the range of other traditional risk factors. For example, a 1 SD change in BMI was associated with an age at first ACS that was 2.2 years earlier. More importantly, at a population level, the impact of these modest changes in age of ACS may have important consequences. Earlier ACS leads to significant lifelong healthcare costs, potential loss of income in younger individuals and higher losses of quality-adjusted years of life.

    Our results highlight the importance of traditional risk factors as well as the contribution of common genetic markers to earlier ACS and may also have clinical implications in understanding the familial risk after premature ACS which is increased nearly twofold in first-degree family members.28 As we have shown, earlier ACS is strongly associated with a high burden of risk factors (ie, male sex, obesity, smoking) with only a modest contribution from common genetic variation. Although a family history of premature MI is defined solely on the basis of age of first MI, our results suggest that a risk-based definition of ‘premature MI’ for probands may be superior in determining which families are truly at higher risk. Whether future genetic studies using novel technologies, such as exome and whole genome sequencing, focusing specifically on unusually early MI cases based on lower traditional risk factor burden, rather than simply on age, will be more effective in identifying novel or rare predisposing variants remains to be seen.29

    Our study has several strengths. This analysis was performed in a highly representative cohort of ACS cases with an age range routinely seen in clinical practice. We also used 30 robust and validated genetic variants from GWAS studies of coronary artery disease to construct our GRS. The strong agreement between our findings and prior studies evaluating traditional risk factors and age of ACS, including the study of Madala et al, which included over 100 000 MI cases, demonstrates the validity of our study sample and the robustness of these associations. Several limitations also deserve mention. First, risk factor data was obtained by self-report and chart review, which may have led to some misclassification. However, since any misclassification is likely to be non-differential it would bias results toward the null and reduce the magnitude of observed associations. Second, certain traditional risk factors such as diabetes and hypercholesterolaemia, which are important risk factors for ACS, were not associated with lower age of first ACS. While this may be due to misclassification or lack of power to detect such an effect, other studies6 ,7 also found no association between younger age of first ACS and these risk factors. It is important to note that this does not indicate that these are not important risk factors for ACS; these results indicate only that the prevalence of these risk factors is not different in young as compared to old ACS cases. Whether refined classification of these risk factors by severity (eg, severe hypercholesterolaemia) would have demonstrated a higher prevalence in earlier ACS requires further study. Our definition of hypercholesterolaemia and hypertension also included statin or hypertension treatment at time of first ACS. Our analyses showed that there was a high degree of colinearity between a history of hypertension and use of medications, such as ACE inhibitors, which precluded using these variables in the final model. Nevertheless, univariate analyses of these variables demonstrated that they were all statistically associated with an older age of first ACS which was likely driving the association between a history of hypertension and age of first ACS. Third, our study only included common genetic variants from MI GWAS published prior to 2012. We did not include more recently discovered common variants or rare variants that could have strengthened the association between the GRS and age. Last, our study sample consisted of a cohort of ACS cases and, therefore, the incidence or relative risks for early ACS in the general population cannot be inferred. However, all risk factors used in our analysis, including the GRS, have been previously associated with increased risk of ACS.

    In summary, a GRS composed of 30 cardiac SNPs was associated with a younger age of first ACS. Although common genetic predisposition modestly contributes to earlier ACS, several traditional risk factors are strongly associated with a markedly lower age of first ACS highlighting the importance of a heavy burden of risk factors as an important contributor to earlier age of first ACS.

    Key messages

    • What is known on this subject?

    • Traditional risk factors (sex, smoking, hypertension, dyslipidemia and diabetes) have been associated with premature acute coronary syndromes (ACS) and a younger age of first ACS, but the impact of a common genetic predisposition on age of first ACS remains unknown.

    • What might this study add?

    • A Genetic Risk Score composed (GRS) of 30 well-established single nucleotide polymorphisms (SNPs) previously associated with myocardial infarction is associated with a younger age of first ACS over and above traditional risk factors. However, traditional risk factors, such as male sex, smoking and obesity are more strongly associated with younger age of first ACS than a GRS.

    • How might this impact on clinical practice?

    • Our study demonstrates that although patients with early ACS have a higher prevalence of common genetic variants that may increase susceptibility to ACS, a high burden of traditional risk factors is likely a more important contributor to earlier ACS highlighting the importance of aggressive risk factor reduction in young adulthood to prevent premature ACS.

    Acknowledgments

    We acknowledge the collaboration of the principal investigators and research personnel of the participating centres in RISCA, in particular Luce Boyer, RN. We also acknowledge the genotyping expertise of Rosalie Frechette and the genotyping platform of the McGill University and Génome Québec Innovation Centre and Katia Desbiens and Michael Gritti for technical help.

    References

    1. ↵
      1. Towfighi A,
      2. Zheng L,
      3. Ovbiagele B
      . Sex-specific trends in midlife coronary heart disease risk and prevalence. Arch Intern Med 2009;169:1762–6.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. D'Agostino RB Sr.,
      2. Vasan RS,
      3. Pencina MJ,
      4. et al
      . General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation 2008;117:743–53.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Cole JH,
      2. Miller JI 3rd.,
      3. Sperling LS,
      4. et al
      . Long-term follow-up of coronary artery disease presenting in young adults. J Am Coll Cardiol 2003;41:521–8.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Berenson GS,
      2. Srinivasan SR,
      3. Bao W,
      4. et al
      . Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. The Bogalusa Heart Study.  N Engl J Med 1998;338:1650–6.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Bähler C,
      2. Gutzwiller F,
      3. Erne P,
      4. et al
      . Lower age at first myocardial infarction in female compared to male smokers. Eur J Prev Cardiol 2012;19:1184–93.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Grundtvig M,
      2. Hagen TP,
      3. German M,
      4. et al
      . Sex-based differences in premature first myocardial infarction caused by smoking: twice as many years lost by women as by men. Eur J Cardiovasc Prev Rehabil 2009;16:174–9.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Madala MC,
      2. Franklin BA,
      3. Chen AY,
      4. et al
      . Obesity and age of first non-ST-segment elevation myocardial infarction. J Am Coll Cardiol 2008;52:979–85.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Khot UN,
      2. Khot MB,
      3. Bajzer CT,
      4. et al
      . Prevalence of conventional risk factors in patients with coronary heart disease. JAMA2003;290:898–904.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Schunkert H,
      2. Konig IR,
      3. Kathiresan S,
      4. et al
      . Large-scale association analysis identifies 13 new susceptibility loci for coronary artery disease. Nat Genet 2011;43:333–8.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Trégouët DA,
      2. König IR,
      3. Erdmann J,
      4. et al
      . Genome-wide haplotype association study identifies the SLC22A3-LPAL2-LPA gene cluster as a risk locus for coronary artery disease. Nat Genet 2009;41:283–5.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Erdmann J,
      2. Großhennig A,
      3. Braund PS,
      4. et al
      . New susceptibility locus for coronary artery disease on chromosome 3q22.3. Nat Genet 2009;41:280–2.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Gudbjartsson DF,
      2. Bjornsdottir US,
      3. Halapi E,
      4. et al
      . Sequence variants affecting eosinophil numbers associate with asthma and myocardial infarction. Nat Genet 2009;41:342–7.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Helgadottir A,
      2. Thorleifsson G,
      3. Manolescu A,
      4. et al
      . A common variant on chromosome 9p21 affects the risk of myocardial infarction. Science 2007;316:1491–3.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Kathiresan S,
      2. Voight BF,
      3. Purcell S,
      4. et al
      . Genome-wide association of early-onset myocardial infarction with single nucleotide polymorphisms and copy number variants. Nat Genet 2009;41:334–41.
      OpenUrlCrossRefPubMed
    15. ↵
      1. McPherson R,
      2. Pertsemlidis A,
      3. Kavaslar N,
      4. et al
      . A common allele on chromosome 9 associated with coronary heart disease. Science 2007;316:1488–91.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Samani NJ,
      2. Erdmann J,
      3. Hall AS,
      4. et al
      . Genomewide association analysis of coronary artery disease. N Engl J Med 2007;357:443–53.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Burton PR,
      2. Clayton DG,
      3. Cardon LR,
      4. et al
      . Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 2007;447:661–78.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Rich JD,
      2. Cannon CP,
      3. Murphy SA,
      4. et al
      . Prior aspirin use and outcomes in acute coronary syndromes. J Am Coll Cardiol 2010;56:1376–85.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Ellis KL,
      2. Pilbrow AP,
      3. Frampton CM,
      4. et al
      . A common variant at chromosome 9P21.3 is associated with age of onset of coronary disease but not subsequent mortality. Circ Cardiovasc Genet 2010;3:286–93.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Meng W,
      2. Hughes AE,
      3. Patterson CC,
      4. et al
      . Chromosome 9p21.3 is associated with early-onset coronary heart disease in the Irish population. Dis Markers 2008;25:81–5.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Cox DR
      . Regression Models and Life-Tables. J R Stat Soc Series B (Methodol) 1972;34:187–220.
      OpenUrl
    22. ↵
      1. Bogaty P,
      2. Boyer L,
      3. Simard S,
      4. et al
      . Clinical utility of C-reactive protein measured at admission, hospital discharge, and 1 month later to predict outcome in patients with acute coronary disease. The RISCA (recurrence and inflammation in the acute coronary syndromes) study. J Am Coll Cardiol 2008;51:2339–46.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Thanassoulis G,
      2. Peloso GM,
      3. Pencina MJ,
      4. et al
      . A genetic risk score is associated with incident cardiovascular disease and coronary artery calcium: the Framingham Heart Study. Circ Cardiovasc Genet 2012;5:113–21.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Patel RS,
      2. Sun YV,
      3. Hartiala J,
      4. et al
      . Association of a genetic risk score with prevalent and incident myocardial infarction in subjects undergoing coronary angiography. Circ Cardiovasc Genet 2012;5:441–9.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Topol EJ,
      2. McCarthy J,
      3. Gabriel S,
      4. et al
      . Single nucleotide polymorphisms in multiple novel thrombospondin genes may be associated with familial premature myocardial infarction. Circulation 2001;104:2641–4.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Bao W,
      2. Srinivasan SR,
      3. Wattigney WA,
      4. et al
      . The relation of parental cardiovascular disease to risk factors in children and young adults. The Bogalusa Heart Study. Circulation 1995;91:365–71.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Lloyd-Jones DM,
      2. Nam BH,
      3. D'Agostino RB Sr.,
      4. et al
      . Parental cardiovascular disease as a risk factor for cardiovascular disease in middle-aged adults: a prospective study of parents and offspring. JAMA 2004;291:2204–11.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Vaidya D,
      2. Yanek LR,
      3. Moy TF,
      4. et al
      . Incidence of coronary artery disease in siblings of patients with premature coronary artery disease: 10 years of follow-up. Am J Cardiol 2007;100:1410–5.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Pencina MJ,
      2. D'Agostino RB Sr.,
      3. D'Agostino RB Jr.,
      4. et al
      . Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008;27:157–72; discussion 207–12.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Contributors All authors contributed equally to this manuscript.

    • Funding This work was supported by Canadian Institute of Health Research grant CIHR MOP-119380 to Dr. Thanassoulis as well as, in part, by CIHR IGO- 86113 to Dr. Pilote. The RISCA cohort was supported in part by the Fonds de la Recherche en Santé du Québec, the Heart and Stroke Foundation of Canada, and unrestricted grants from Merck Frosst Canada and Pfizer Canada.

    • Competing interests GT is supported by a FRQS Chercheur Boursier Clinicien Salary Award. None of the authors have any relationship with private industry relevant to this manuscript.

    • Ethics approval Each hospital committee on human research approved the study.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    Linked Articles