Abstract
Background:
Current cardiovascular risk assessment guidelines incorporate judicious use of C-reactive protein (CRP), carotid intima-media thickness (CIMT), and coronary artery calcium (CAC) in selected populations and describe threshold levels for higher and lower cardiovascular risk for each of the three risk refinement tests. However, the effect of these suggested thresholds of relative risk on absolute global risk remains uncertain.
Methods:
Systematic permutation of risk factors provided 10-year risk estimates using the Framingham risk score, equations derived from the Multi-Ethnic Study of Atherosclerosis (MESA) and the Atherosclerosis Risk in Communities (ARIC) study, and the Reynolds risk score. Low-, high-, and very-high-risk values of CAC, CIMT, and hsCRP were defined as: 0, 100, 400 Agatston units; 25th percentile without plaque, 75th percentile without plaque, 75th percentile with plaque; and 1.0, 3.0, 7.0 mg/L.
Results:
Incorporation of low-, high-, and very-high-risk CAC values using the MESA risk score resulted in greater changes in absolute risk from the Framingham risk score than the addition of either CIMT or hsCRP values using the ARIC or Reynolds risk scores.
Conclusions:
Although certain values of CAC, CIMT, and hsCRP have been similarly designated as low, high, or very-high risk, incorporation of these thresholds into validated risk equations yielded substantially different levels of absolute cardiovascular risk. Use of available risk equations may be advisable to calculate absolute risk rather than relying on risk-marker thresholds derived from relative risk estimates.
Keywords: Cardiovascular risk, Risk equations, C-reactive protein, Carotid intima-media thickness, Coronary artery calcium
1. Introduction
The National Cholesterol Education Program’s Adult Treatment Panel III (ATP-III) report recommends assessing the patient’s 10-year absolute risk of coronary heart disease (CHD) events based on readily available clinical risk factors, and matching the intensity of CHD prevention efforts to prognosis for CHD events [1]. The ATP-III approach is based on ‘hard’ CHD events: CHD death and non-fatal myocardial infarction (MI). Accordingly, ATP-III used a set of risk-prediction equations derived from the Framingham cohort to inform risk categories for 10-year incidence of hard CHD events: low risk (<10%), intermediate risk (10–20%), and high risk (>20%).
Increasingly, physicians have access to more sophisticated tests in addition to those clinical risk factors included in the Framingham risk equation, fueling interest in augmenting the ATP-III risk categories with novel risk markers to refine CHD prognosis. For example, coronary artery calcium (CAC) by computed tomography provides a direct assessment of subclinical CHD, and has been shown to accurately predict CHD events [2]. Atherosclerosis outside the coronary system, as quantified by carotid intima-media thickness (CIMT) and the presence of carotid plaque, also predicts CHD events [3]. In addition, novel biomarkers, such as high-sensitivity C-reactive protein (hsCRP), may offer a low-cost and low-intensity improvement in CHD risk assessment [4].
Subsequent clinical guidelines incorporate judicious use of CAC, CIMT, or hsCRP in selected populations and describe threshold levels for higher and lower cardiovascular risk [5]. Discrete thresholds for novel risk markers offer a practical simplification of continuous variables and facilitate clinical decision-making. However, overreliance on thresholds for novel risk markers outside the context of the ATP-III risk factors may divert attention from the fundamental goal of CHD risk assessment: evaluation of absolute risk. For example, a lower-risk hsCRP value alone, defined as a level below 1.0 mg/L, does not imply low risk in absolute terms (i.e. 10-year hard CHD risk below 10% per ATP-III); nor does a higher-risk hsCRP alone imply high absolute risk. Rather, clinical management recommendations are derived from near-term, absolute risk as defined by comprehensive assessment of the ATP-III risk factors and, in selected individuals, further risk-refinement testing. Fortunately, equations have been derived from large prospective cohort studies that calculate absolute risk by ATP-III risk factors as well as novel risk markers: the Multi-Ethnic Study of Atherosclerosis (MESA) [6] incorporates CAC; the Atherosclerosis Risk in Communities (ARIC) study [7,8] incorporates CIMT; and the Reynolds risk score, derived from the Women’s Health Study [9] and the Physician’s Health Study [10], incorporates hsCRP.
The purpose of the present study is to examine differences in absolute 10-year cardiovascular risk between alternative risk-refinement equations by systematically varying the clinical and laboratory variables to those equations across a full range of plausible values. Our goal is to determine the extent to which differences in novel risk markers could impact CHD prognosis and clinical decision-making. Systematically varying risk-equation inputs was used to clarify the properties of the Framingham 10-year risk equation, revealing the paucity of combinations that yield a high risk score among women [11]. We applied similar methods to compare the Framingham CHD risk score to the absolute risks associated with suggested risk thresholds for CAC by MESA, CIMT by ARIC, and hsCRP by Reynolds risk equations. Because this approach does not depend on sampling individuals from a specific population, it evaluates the effects of all permutations of individual risk markers and how aggregate risk-marker burden impacts 10-year CHD risk predictions. We hypothesized that, for a given combination of ATP-III risk factors, low-, high-, and very-high-risk CAC, CIMT, and hsCRP values would yield different results in terms of absolute risk.
2. Methods
ATP-III clinical risk factors were systematically permuted for all covariate values shown in Supplementary Table 1. For every combination of clinical risk factors, one Framingham risk score, three MESA risk scores (low/high/very-high-risk CAC), three ARIC risk scores (low/high/very-high-risk CIMT), and three Reynolds risk scores (low/high/very-high-risk hsCRP) were calculated. As recommended by the 2006 American Heart Association consensus statement, a CAC score of 0 Agatston units was defined as low risk; CAC scores of 100 and 400 Agatston units were considered high risk and very high risk, respectively [2]. As recommended by the 2008 American Society of Echocardiogram/Society of Vascular Medicine guidelines [3], a CIMT value at the 25th percentile with no evidence of carotid plaque was defined as low risk; CIMT values at the 75th percentile without and with carotid plaque were defined as high and very high risk. CIMT thresholds adjusted for age and sex (Supplementary Table 2). As recommended by the 2003 American Heart Association/Centers for Disease Control consensus statement [4], hsCRP values of 1.0 and 3.0 mg/L were defined as low and high risk, respectively. A hsCRP value of 7.0 mg/L was defined as very high risk; 7.0 mg/L represents the 75th percentile of hsCRP in the Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin trial (JUPITER), a clinical trial of rosuvastatin among individuals free of cardiovascular disease at baseline with elevated hsCRP [12]. Baseline characteristics of the derivation cohorts for all risk scores are provided in Supplementary Table 3 [7,9,10,13,14]. All risk equations were previously validated across the range of inputs used in our analysis.
We calculated absolute 10-year risk estimates of hard CHD for the Framingham, MESA, and ARIC risk equations. The Reynolds score did not attempt to evaluate hard CHD risk, making it more challenging to evaluate in the ATP-III framework and also making it difficult to compare with the other three. While hard CHD events are included, the Reynolds risk score additionally incorporates non-CHD death and coronary revascularization, going beyond hard CHD events, and ischemic stroke, going beyond the coronary system. Nevertheless, we thought it would be helpful to include the Reynolds score in this analysis because its ease of use and low cost enhance its clinical potential. In sensitivity analyses, we compared the Reynolds risk score to two Framingham risk equations that calculate more inclusive composite endpoints: the ‘all’ CHD Framingham risk score, which incorporates coronary insufficiency and angina; and the ‘general’ cardiovascular disease (CVD) Framingham risk score, which additionally incorporates hemorrhagic stroke, peripheral artery disease, and heart failure.
Across all risk-factor combinations, graphical summaries were produced based on scatterplot smoothing splines for (1) absolute risk estimates from all risk-refinement equations vs. the Framingham risk score and (2) differences in absolute risk from all risk-refinement equations and the Framingham risk score vs. the Framingham risk score. At levels of Framingham risk between 10% and 20%, we calculated medians and the inter-quartile ranges of absolute risk differences between very-high-, high-, and low-risk groups across all risk-factor combinations. All analyses were completed using R 2.15.1 (R Development Core Team, Vienna, Austria) including the knitr extension package (Xie, 2012). The Technical Appendix provides all R commands used to define the risk equations [6–10,14] and to generate the data.
3. Results
Absolute risk scores were calculated by systematically varying inputs for the Framingham, MESA, ARIC, and Reynolds risk equations (Fig. 1). Incorporation of suggested low-, high-, and very-high-risk CAC values using the MESA risk score resulted in greater changes in absolute risk from the Framingham risk score than the addition of either low-, high-, and very-high-risk CIMT or hsCRP values using the ARIC or Reynolds risk scores, respectively (Fig. 2). Among permutations of risk factors yielding Framingham risk scores of 10—20% in men, the absolute difference between MESA risk scores generated from high- and low-risk CAC values ranged 11.8—14.6% in men (Table 1). Smaller differences in absolute risk were observed for ARIC risk scores derived from high- and low-risk CIMT values (2.5—4.4%) or Reynolds risk scores calculated using high- and low-risk hsCRP values (1.0—1.9%) (Table 1). Similar findings were observed among women and for absolute differences between risk scores derived from very-high-risk and low-risk values of CAC, CIMT, and hsCRP (Table 1). Sensitivity analyses comparing the Reynolds risk score with the more inclusive Framingham all CHD and general CVD risk equations did not reveal substantial differences in changes in absolute risk.
Fig. 1.
10-year absolute risk calculated using the Framingham and MESA (A), ARIC (B), or Reynolds risk equations (C) in men and women. For permutations resulting in a given Framingham risk score, smoothing splines for the Reynolds risk scores using low, high, and very high risk CRP values were computed, keeping other inputs constant. Similarly, smoothing splines of the ARIC risk scores using low, high, and very high risk CIMT values and MESA risk scores using low, high, and very high risk CAC values were calculated.
Fig. 2.
10-year absolute risk differences between Framingham and MESA (A), ARIC (B), or Reynolds (C) risk equations in men and women.
Table 1.
Absolute risk difference (%) between risk groups for the Reynolds, ARIC, and MESA risk equations by Framingham risk (10–20%); median (25th, 75th percentile) are provided across all possible inputs (see Supplementary Table 1) a. Among men, b. Among women.
| Risk | Reynolds |
ARIC |
MESA |
|||
|---|---|---|---|---|---|---|
| High vs. low | Very high vs. low | High vs. low | Very high vs. low | High vs. low | Very high vs. low | |
| a. Among men | ||||||
| 10% | 1.0 (0.8, 1.4) | 1.8 (1.4, 2.5) | 2.6 (1.0, 4.6) | 5.2 (2.8, 8.7) | 14.2 (10.2, 18.2) | 17.8 (13.3, 23.0) |
| 11% | 1.1 (0.8, 1.4) | 2.0 (1.4, 2.6) | 2.5 (0.7, 5.2) | 5.9 (2.8, 9.0) | 14.1 (9.7, 19.5) | 17.8 (13.9, 24.1) |
| 12% | 1.2 (0.9, 1.5) | 2.2 (1.6, 2.8) | 3.0 (0.5, 6.1) | 6.4 (3.5, 10.5) | 13.8 (8.7, 18.0) | 18.0 (12.5, 22.8) |
| 13% | 1.3 (1.0, 1.7) | 2.4 (1.8, 3.1) | 3.1 (0.6, 6.2) | 7.3 (3.5, 10.7) | 14.4 (9.0, 18.8) | 19.0 (13.2, 23.7) |
| 14% | 1.4 (1.0, 1.8) | 2.6 (1.9, 3.4) | 3.0 (0.7, 6.5) | 6.9 (3.6, 11.4) | 14.1 (9.5, 19.2) | 18.6 (13.6, 23.8) |
| 15% | 1.4 (1.1, 2.0) | 2.6 (2.1, 3.6) | 3.5 (0.7, 7.4) | 7.8 (3.8, 12.1) | 14.6 (8.2, 19.1) | 18.6 (11.7, 23.6) |
| 16% | 1.6 (1.1, 2.2) | 3.0 (2.1, 4.0) | 3.6 (1.1, 6.8) | 7.7 (4.0, 12.0) | 13.7 (8.9, 19.7) | 19.5 (12.8, 25.0) |
| 17% | 1.7 (1.2, 2.3) | 3.2 (2.3, 4.2) | 4.2 (1.0, 7.3) | 8.7 (4.5, 13.6) | 12.9 (8.7, 19.2) | 18.5 (12.4, 24.0) |
| 18% | 1.8 (1.3, 2.4) | 3.3 (2.4, 4.4) | 4.0 (1.5, 7.5) | 8.7 (5.2, 12.8) | 13.0 (6.6, 19.9) | 18.7 (9.5, 25.0) |
| 19% | 1.9 (1.4, 2.5) | 3.4 (2.5, 4.5) | 4.4 (1.0, 7.6) | 9.0 (5.8, 13.9) | 13.2 (8.0, 21.0) | 18.9 (11.4, 26.3) |
| 20% | 1.9 (1.4, 2.5) | 3.5 (2.6, 4.6) | 4.2 (1.7, 7.7) | 9.0 (6.0, 13.6) | 11.8 (7.1, 19.2) | 16.7 (10.2, 24.8) |
| b. Among women | ||||||
| 10% | 2.1 (1.5, 2.7) | 4.0 (2.9, 5.1) | 1.2 (−0.4, 3.1) | 7.5 (4.4, 9.9) | 8.7 (5.7, 13.9) | 14.7 (9.7, 22.9) |
| 11% | 2.4 (1.7, 3.2) | 4.5 (3.2, 6.0) | 1.5 (−0.5, 3.4) | 8.1 (4.6, 11.0) | 9.2 (5.8, 12.4) | 15.5 (10.0, 20.7) |
| 12% | 2.5 (1.8, 3.1) | 4.7 (3.5, 5.8) | 1.4 (0.1, 2.7) | 7.9 (5.9, 10.3) | 9.3 (6.0, 15.2) | 15.5 (10.3, 24.8) |
| 13% | 2.7 (2.0, 3.7) | 5.1 (3.7, 6.9) | 1.5 (−0.1, 3.6) | 8.6 (6.0, 11.3) | 9.2 (6.4, 12.4) | 15.5 (11.0, 20.7) |
| 14% | 2.9 (2.1, 3.6) | 5.4 (4.0, 6.8) | 2.0 (−0.7, 3.4) | 9.3 (7.0, 12.3) | 9.6 (5.6, 14.2) | 15.9 (10.2, 23.8) |
| 15% | 2.8 (2.0, 3.8) | 5.3 (3.8, 7.1) | 1.7 (0.0, 4.4) | 9.6 (6.7, 12.5) | 10.2 (7.0, 14.5) | 16.9 (12.1, 24.5) |
| 16% | 3.3 (2.5, 4.0) | 6.2 (4.7, 7.5) | 2.0 (−0.8, 3.6) | 10.4 (7.3, 12.7) | 9.8 (6.1, 15.3) | 16.4 (11.2, 25.0) |
| 17% | 3.3 (2.3, 4.2) | 6.2 (4.3, 7.9) | 2.1 (0.2, 3.9) | 10.9 (8.7, 13.6) | 9.8 (6.3, 13.7) | 16.4 (11.5, 22.6) |
| 18% | 3.2 (2.6, 4.4) | 6.1 (4.9, 8.2) | 2.2 (−0.1, 3.4) | 10.5 (7.8, 14.0) | 10.6 (6.5, 17.7) | 17.7 (11.8, 28.7) |
| 19% | 3.9 (2.8, 4.5) | 7.3 (5.4, 8.4) | 2.1 (0.5, 3.9) | 11.1 (10.0, 13.3) | 10.1 (6.9, 14.8) | 16.9 (12.6, 24.0) |
| 20% | 3.6 (2.6, 4.9) | 6.8 (4.8, 9.1) | 2.6 (0.8, 4.0) | 12.1 (10.1, 14.0) | 9.8 (6.9, 13.9) | 16.4 (12.5, 22.7) |
4. Discussion
To our knowledge, this is the first study to systematically compare the absolute cardiovascular risk estimates generated by four important cardiovascular risk equations. In general, our results demonstrated that the absolute difference in risk for CIMT was more than double that of hsCRP, and in turn, CAC was approximately triple that of CIMT. This affirms the intuitive concept that knowledge of subclinical atherosclerosis in the coronary bed itself provides the greatest prognostic information for CHD above and beyond clinical risk factors.
While these findings could be intuited from prior analyses of CAC [15–21], CIMT [7,19,20], and hsCRP [9,10,19,22], it is important to note that few studies of these markers explicitly detail incremental prognostic value in terms of change in absolute risk. Association, calibration, discrimination, and reclassification are typically described using multivariable-adjusted relative risk estimates, Hosmer-Lemeshow goodness-of-fit tests, area under the curve, and net reclassification improvement, respectively [22,23]; however, none of these summary statistics are readily translated to the clinical recommendation to stage risk by 10-year absolute risk of hard CHD events. Our analysis, particularly the graphical representation of global absolute risk scores, provides a clinically valuable estimate of the absolute risk based on a particular starting Framingham risk score and the addition of specified risk-refinement test results. In addition, since available summary statistics are calculated from cohort study data, they are limited to the characteristics of the cohort, whereas our empirical analysis was performed across the plausible range for covariate values.
Post-ATP-III guidelines consider the assessment of hsCRP, CIMT, or CAC to be reasonable to refine cardiovascular risk in the setting of an intermediate 10-year Framingham risk score of 10—20% [5]. However, given the different incremental prognostic value of each test, it is not at all clear that identical Framingham risk ‘entry criteria’ should apply to all three. For example, on average, when the Framingham risk score exceeds 15%, addition of a very-high-risk CRP value in men results in an absolute cardiovascular event rate above 20% according to the Reynolds risk equation (Fig. 1). On the other hand, a very-high-risk CAC value in men yields an absolute cardiovascular event rate above 20% according to the MESA risk equation, while the Framingham risk score is less than 5% (Fig. 1). In other words, upgrading risk for individuals to the ATP-III’s high-risk category (>20%) would be uncommon for high CRP values unless the Framingham risk score exceeds 15%. This suggests that, when CRP is used to identify high-risk patients, it may be best to restrict testing to patients with intermediate-high Framingham risk scores. On the other hand, high CAC upgrades risk when Framingham risk scores are below 5%, suggesting that CAC may be useful to refine cardiovascular risk even among patients at low Framingham risk. In theory, a similar calculus applies to lower risk scores for downgrading risk.
Our analysis is limited by comparing cardiovascular risk scores derived from different cohorts. Among different populations, there may be different associations between risk factors and outcomes. Thus, differences in absolute risk between risk scores may be partially due to these different associations. However, we believe the present analysis is nonetheless insightful. In particular, absolute risk difference within each of the three risk refinement equations using accepted higher and lower risk values of CAC, CIMT, or hsCRP represents a within-cohort ‘dynamic range’ of the absolute risk resulting from the addition of a risk-refinement measure and may represent best-case scenarios for these tests. Studies following the Physician’s Health Study and the Women’s Health Study have reported an attenuated effect of hsCRP on incremental risk prediction [24]; similarly, analyses conducted in cohorts other than the ARIC study have demonstrated a smaller impact of CIMT when added to the ATP-III risk factors [20,25]. After accrual of a sufficient number of incident cardiovascular events, data from MESA will permit the development of robust prediction equations of absolute risk, and a direct comparison between equations incorporating CAC, CIMT, and hsCRP can be performed utilizing a single, well-characterized cohort. Until that time, the MESA, ARIC, and Reynolds risk equations remain the only validated, clinically available cardiovascular risk equations incorporating risk-refinement measures.
A second limitation of the present analysis is that the Reynolds risk equation is based on a broad composite endpoint of myocardial infarction, stroke, coronary revascularization, and any cardiovascular death, compared to the other three risk equations whose endpoints are restricted to MI and fatal CHD. 10-year hard CHD Reynolds risk equations are not currently available. While this may limit direct comparisons between the risk equations, the more diffuse endpoint should bias the magnitude of absolute risk differences upward. Notably, despite the broader composite end-point, absolute risk differences in the Reynolds risk score remain modest with incorporation of low-, high-, or very-high-risk hsCRP values. In addition, sensitivity analyses comparing the Reynolds risk score to more inclusive all CHD and general CVD Framingham risk equations did not alter our results.
Current guidelines provide lower and higher risk thresholds for CAC, CIMT, and hsCRP. Although these values represent thresholds for relative risk, in clinical practice they may be misconstrued as thresholds for absolute risk. This has significant implications for primary preventive care. For example, assuming high near-term absolute risk from a hsCRP above the ‘high’ risk threshold of 3.0 mg/L may overestimate absolute risk, resulting in overtreatment. Conversely, assuming low near-term absolute risk from a hsCRP below the ‘low’ risk threshold of 1.0 mg/L may underestimate absolute risk, resulting in undertreatment. While similar misclassifications may occur using ‘high’ and ‘low’ risk thresholds for CAC, they appear to occur less frequently. As shown in Fig. 1, a ‘high’ risk CAC threshold of 100 AU more commonly results in an absolute risk that meets the threshold of high absolute risk, i.e. >20%. For risk-refinement tests, use of available clinical risk equations may be advisable to determine absolute cardiovascular risk.
Assessment of absolute risk, informed largely but not exclusively by risk equations, has been proposed as the first step of management in contemporary preventive cardiovascular guidelines [26,27] because risk equations have proven more accurate in prognosticating risk compared to an ad-hoc or unifactor-based approach or risk factor-counting methods [23,28]. Although there are very limited data to suggest that using risk equations improves outcomes, the rationale of matching intensity of therapy to degree of risk is supported by pharmacotherapeutic trials demonstrating greater absolute risk reduction for CHD and stroke among those at higher baseline risk than lower-risk persons [29]. As summarized by the NCEP ATP II, “The intensity of treatment of the individual patient depends on the patient’s risk status. Risk is stratified according to the likelihood of developing CHD in the relatively near future. Those at higher risk for developing CHD in the short term should receive more aggressive intervention than patients at lower risk [30].” Finally, despite recommendations to incorporate the use of risk factor equations, underutilization persists in routine primary care, likely due to the added time and inconvenience required for their calculation. Efforts to automate their calculation through programs leveraging electronic medical records may facilitate more widespread use.
Supplementary Material
Acknowledgments
Dr. Dunbar was supported by K23HL091130. The authors thank Dr. Vijay Nambi and the Atherosclerosis Risk in Communities Coordinating Center for providing the risk equation incorporating carotid intima-media thickness.
Footnotes
Competing interests
The authors declare that they have no competing interests.
Financial disclosures
None.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.atherosclerosis.2012.12.025.
References
- [1].Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–421. [PubMed] [Google Scholar]
- [2].Budoff MJ, Achenbach S, Blumenthal RS, et al. Assessment of coronary artery disease by cardiac computed tomography: a scientific statement from the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology. Circulation 2006;114:1761–91. [DOI] [PubMed] [Google Scholar]
- [3].Stein JH, Korcarz CE, Hurst RT, et al. Use of carotid ultrasound to identify subclinical vascular disease and evaluate cardiovascular disease risk: a consensus statement from the American Society of Echocardiography Carotid Intima-Media Thickness Task Force. Endorsed by the Society for Vascular Medicine. J Am Soc Echocardiogr 2008;21:93–111 [quiz 189–90]. [DOI] [PubMed] [Google Scholar]
- [4].Pearson TA, Mensah GA, Alexander RW, et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 2003;107:499–511. [DOI] [PubMed] [Google Scholar]
- [5].Greenland P, Alpert JS, Beller GA, et al. ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2010;122:e584–636. [DOI] [PubMed] [Google Scholar]
- [6].McClelland RL, Nasir K, Budoff M, Blumenthal RS, Kronmal RA. Arterial age as a function of coronary artery calcium (from the Multi-Ethnic Study of Atherosclerosis [MESA]). Am J Cardiol 2009;103:59–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Nambi V, Chambless L, Folsom AR, et al. Carotid intima-media thickness and presence or absence of plaque improves prediction of coronary heart disease risk: the ARIC (Atherosclerosis Risk in Communities) study. J Am Coll Cardiol 2010;55:1600–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].ARIC. ARIC coronary heart disease risk calculator that includes carotid ultrasound information; 2010. [Google Scholar]
- [9].Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds Risk Score. JAMA 2007;297:611–9. [DOI] [PubMed] [Google Scholar]
- [10].Ridker PM, Paynter NP, Rifai N, Gaziano JM, Cook NR. C-reactive protein and parental history improve global cardiovascular risk prediction: the Reynolds Risk Score for men. Circulation 2008;118:2243–51 [2244p following 2251]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Cavanaugh-Hussey MW, Berry JD, Lloyd-Jones DM. Who exceeds ATP-III risk thresholds? Systematic examination of the effect of varying age and risk factor levels in the ATP-III risk assessment tool. Prev Med 2008;47:619–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Ridker PM, Danielson E, Fonseca FA, et al. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med 2008;359:2195–207. [DOI] [PubMed] [Google Scholar]
- [13].Goff DC Jr, Bertoni AG, Kramer H, et al. Dyslipidemia prevalence, treatment, and control in the Multi-Ethnic Study of Atherosclerosis (MESA): gender, ethnicity, and coronary artery calcium. Circulation 2006;113:647–56. [DOI] [PubMed] [Google Scholar]
- [14].Wilson PW, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837–47. [DOI] [PubMed] [Google Scholar]
- [15].Arad Y, Goodman KJ, Roth M, Newstein D, Guerci AD. Coronary calcification, coronary disease risk factors, C-reactive protein, and atherosclerotic cardiovascular disease events: the St. Francis Heart Study. J Am Coll Cardiol 2005;46:158–65. [DOI] [PubMed] [Google Scholar]
- [16].Kondos GT, Hoff JA, Sevrukov A, et al. Electron-beam tomography coronary artery calcium and cardiac events: a 37-month follow-up of 5635 initially asymptomatic low- to intermediate-risk adults. Circulation 2003;107:2571–6. [DOI] [PubMed] [Google Scholar]
- [17].LaMonte MJ, FitzGerald SJ, Church TS, et al. Coronary artery calcium score and coronary heart disease events in a large cohort of asymptomatic men and women. Am J Epidemiol 2005;162:421–9. [DOI] [PubMed] [Google Scholar]
- [18].Preis SR, Hwang SJ, Fox CS, et al. Eligibility of individuals with subclinical coronary artery calcium and intermediate coronary heart disease risk for reclassification (from the Framingham Heart Study). Am J Cardiol 2009;103:1710–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Yeboah J, McClelland RL, Polonsky TS, et al. Comparison of novel risk markers for improvement in cardiovascular risk assessment in intermediate-risk individuals. JAMA 2012;308:788–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Folsom AR, Kronmal RA, Detrano RC, et al. Coronary artery calcification compared with carotid intima-media thickness in the prediction of cardiovascular disease incidence: the Multi-Ethnic Study of Atherosclerosis (MESA). Arch Intern Med 2008;168:1333–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Polonsky TS, McClelland RL, Jorgensen NW, et al. Coronary artery calcium score and risk classification for coronary heart disease prediction. JAMA 2010;303:1610–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Cook NR, Paynter NP, Eaton CB, et al. Comparison of the Framingham and Reynolds Risk scores for global cardiovascular risk prediction in the multiethnic Women’s Health Initiative. Circulation 2012;125:1748–56. S1711–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Lloyd-Jones DM. Cardiovascular risk prediction: basic concepts, current status, and future directions. Circulation 2010;121:1768–77. [DOI] [PubMed] [Google Scholar]
- [24].Blaha MJ, Budoff MJ, DeFilippis AP, et al. Associations between C-reactive protein, coronary artery calcium, and cardiovascular events: implications for the JUPITER population from MESA, a population-based cohort study. Lancet 2011;378:684–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Polak JF, Pencina MJ, Pencina KM, O’Donnell CJ, Wolf PA, D’Agostino RB Sr. Carotid-wall intima-media thickness and cardiovascular events. N Engl J Med 2011;365:213–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110:227–39. [DOI] [PubMed] [Google Scholar]
- [27].Mancia G, De Backer G, Dominiczak A, et al. 2007 Guidelines for the management of arterial hypertension: The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J 2007;28:1462–536. [DOI] [PubMed] [Google Scholar]
- [28].Cooney MT, Dudina AL, Graham IM. Value and limitations of existing scores for the assessment of cardiovascular risk: a review for clinicians. J Am Coll Cardiol 2009;54:1209–27. [DOI] [PubMed] [Google Scholar]
- [29].Jackson R, Lawes CM, Bennett DA, Milne RJ, Rodgers A. Treatment with drugs to lower blood pressure and blood cholesterol based on an individual’s absolute cardiovascular risk. Lancet 2005;365:434–41. [DOI] [PubMed] [Google Scholar]
- [30].Summary of the second report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II). JAMA 1993;269:3015–23. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.