Author + information
- Received May 10, 2005
- Revision received July 1, 2005
- Accepted July 19, 2005
- Published online November 15, 2005.
- ↵⁎Reprint requests and correspondence:
Dr. Paolo Raggi, 1430 Tulane Avenue, SL-48, New Orleans, Louisiana 70112
Objectives The purpose of this study was to investigate the classification of cardiovascular risk in young individuals and women according to the National Cholesterol Education Program (NCEP) guidelines across a continuum of coronary calcium scores (CCS).
Background Current NCEP guidelines might underestimate cardiovascular risk in young individuals and women.
Methods The study population consisted of 1,611 asymptomatic individuals (67% men, mean age: 53 ± 10 years) who presented to a single electron beam tomography facility for coronary artery calcium screening. Participants were categorized into low-risk (n = 738, 46%), intermediate-risk (n = 583, 36%), moderately high-risk (n = 263, 16%), and high-risk (n = 27, 2%) according to the NCEP Panel III guidelines.
Results Absence of calcium, CCS of 0 to 99 (mild), 100 to 399 (moderate), and ≥400 (severe), was observed in 572 (35%), 707 (44%), 192 (12%), and 140 (9%) of the patients, respectively. A high CCS percentile (≥75th percentile) was present in 426 (26%) individuals. Overall, 59% and 78% of participants with CCS ≥400 and CCS ≥75th percentile were not identified as high risk and candidates for pharmacotherapy on the basis of NCEP categories. Furthermore, women as well as young individuals were less likely to be considered candidates for pharmacotherapy compared with men and older individuals in each CCS category.
Conclusions The NCEP guidelines seem to underestimate cardiovascular risk in young asymptomatic individuals and women. For these individuals, assessment of plaque burden might provide incremental value to global risk assessment.
Coronary heart disease (CHD) affects over one million Americans annually and is the leading cause of death in the U.S., having caused nearly one-half of a million deaths among the adult population in 2001 (1). As many as one-half of the first coronary events (including sudden cardiac death) occur in asymptomatic people (2), which underscores the importance of detecting individuals at risk before an initial event to implement primary prevention strategies (3).
The primary recommendation of several advisory bodies is that all adults undergo an office-based assessment as the initial step to identify those at higher risk for CHD. One approach adopted by the National Cholesterol Education Program Panel III (NCEP-III) is to apply a modification of the risk prediction algorithm derived from the Framingham Heart Study to estimate a person’s 10-year risk for developing CHD (3,4). The NCEP-III guidelines suggest using the Framingham risk categories to target low-density lipoprotein (LDL) as the primary goal of preventive treatment (4,5). A potential limitation of the NCEP-III guidelines is the underestimation of risk in young individuals and women, as highlighted in previous reports (6,7).
Identification of subclinical atherosclerosis and coronary plaque burden might help to improve identification of high-risk individuals. Evidence exists that coronary calcium scores (CCS) quantified by electron beam tomography (EBT) are a reasonable surrogate for coronary atherosclerosis burden in adults (8–10). Furthermore, in recent years, a number of studies have demonstrated the independent prognostic value of increasing CCS in predicting future CHD events in asymptomatic individuals (11–17).
Several reports have shown that CCS ≥400 and CCS ≥75th percentile for age-gender are associated with a significant CHD risk (11,15,16). Experts from the NCEP-III panel have recommended quantification of CCS as an option for advanced risk assessment in the intermediate-risk categories and indicated that a high CCS (e.g., ≥75th percentile for age and gender) provides a rationale for intensified medical therapy.
Our study was designed to investigate the eligibility for pharmacotherapy according to NCEP-III guidelines among asymptomatic individuals across a spectrum of CCS. Secondarily, we evaluated whether a difference in risk stratification by NCEP-III guidelines existed across age and gender in individuals with similar atherosclerotic burden.
This is a cross-sectional study on a consecutive sample of 2,046 physician-referred patients who presented to a single EBT scanning facility. We excluded 404 subjects who reported a history of prior myocardial infarction, or coronary and/or peripheral arterial revascularization, current symptoms suggestive of angina, and individuals with diabetes, because they are considered CHD-equivalent irrespective of baseline risk. Additionally, 31 individuals reporting use of cholesterol-lowering medications were also excluded from the analyses. This was done because we used fasting lipid levels to calculate a Framingham risk score and the use of lipid-lowering agents could have interfered with our risk assessment. The study was approved by the local institutional review board and received a waiver of patient consent.
Risk-factor assessment and calculation of 10-year CHD risk
All individuals provided details of their demographics, medical history, medication usage, and current symptoms (by means of questionnaire offered at the time of first scanning). Hypertension was defined by current or recommended use of antihypertensive medications for blood pressure control. Smoking was defined as current tobacco usage. Family history of CHD was defined as CHD in first-degree male relatives ≤55 years of age or female relatives ≤65 years old. Body mass index was calculated as weight (kg)/height (m)2. Young age was defined for men as age <55 years and, for women, <65 years.
Fasting total cholesterol and triglycerides were determined with enzymatic methods on samples drawn on the same day as the EBT. High-density lipoprotein (HDL) cholesterol was measured after precipitation of apolipoprotein B containing particles with phosphotungstate. Low-density lipoprotein cholesterol was calculated with the Friedewald equation.
We categorized participants into low-risk (0 to 1 risk factors), intermediate-risk (≥2 risk factors but <10% risk of CHD at 10 years), moderately high-risk (≥2 risk factors and 10% to 20% risk of CHD in10 years), or high-risk (≥2 risk factors and >20% risk of CHD in 10 years) groups, according to the most recent NCEP guidelines (5). Risk factors considered were cigarette smoking, hypertension (blood pressure ≥140/90 mm Hg or receiving antihypertensive medication), low HDL cholesterol (<40 mg/dl), family history of premature CHD (CHD in male first-degree relative <55 years of age; CHD in female first-degree relative <65 years of age), and age (men ≥45 years; women ≥55 years). The LDL cholesterol cutoffs recommended by the NCEP III guidelines for initiation of lipid-lowering therapy in each risk category are as follows: low-risk (≥190 mg/dl), intermediate-risk (≥160 mg/dl), moderately high-risk (≥130 mg/dl), and high-risk (≥100 mg/dl) (4).
Each patient underwent EBT scanning with a GE Imatron C-150 scanner (GE/Imatron, South San Francisco, California). Coronary arteries were imaged with rapid acquisition of 30 to 40 contiguous slices with a thickness of 3 mm (26-cm2field of view) during end-diastole. Image acquisition was electrocardiographically triggered and occurred during a single 30- to 35-s breath hold. Coronary calcium score was quantified with the previously described Agatston scoring method (18). Calcium was considered present in a coronary artery when a density of >130 Hounsfield units was detected in >3 contiguous pixels (>1.03 mm3) overlying that coronary artery. The CCS was computed from the product of the attenuation factor and the area of a calcification. A total CCS was computed as the sum of all scores from all individual lesions in each coronary artery.
To facilitate data interpretation, the CCS was classified into the following categories: 0, 1.0 to 99, 100 to 399, and ≥400 (no identifiable plaque, mild, moderate, and severe atherosclerotic plaque burden, respectively). These categories of CCS have been used to differentiate between very low, moderate, moderately high, and high cardiovascular risk (19). A CCS ≥75th percentile for age- and gender-matched individuals was also considered as an indicator of high-risk, as suggested in prior reports (11,14,15). The cutoffs for CCS ≥75th percentile were derived from a recent study demonstrating the normal distribution of CCS in 12,936 asymptomatic individuals (20).
Continuous variables are expressed as mean ± SD. Descriptive statistics were used to summarize patient characteristics. Tests of significance between groups were determined with Student ttest for continuous variables and chi-square analysis for categorical variables. The association between increasing 10-year CHD risk categories and the extent of subclinical atherosclerosis as measured by CCS was examined with logistic regression. For each model, the reference group comprised negative scores (CCS = 0). All statistical analyses were performed with Stata version 8.0 (Stata Corp., Austin, Texas). The level of significance was set at p < 0.05 (two-tailed).
The final study population consisted of 1,611 asymptomatic men and women. Demographic and clinical characteristics of the study population are shown in Table 1.The mean age of the study population (67% men) was 53 ± 10 years, and the majority of individuals were relatively young (men <55 years and women <65 years). Over one-half (54%) of the participants had ≥2 CHD risk factors.
Coronary artery calcium scores
Table 2shows the distribution of absolute CCS among men and women. The median (interquartile range) CCS was 6 (0 to 67). Overall calcium scores of 0, 1 to 99, 100 to 399, and ≥400 were observed in 35%, 44%, 12%, and 9% of the study subjects, respectively. Advanced CCS (≥75th percentile) was present in 426 (26%) individuals. As compared with women, men had a higher prevalence of any CCS (72% vs. 50%, p < 0.00001) as well as moderate-to-severe CCS (25% vs. 13%, p < 0.0001), respectively. This trend was observed in both young and older participants.
Ten-year CHD risk
The mean overall 10-year CHD risk was 5.4 ± 5.2%, with a higher risk observed in men than in women (6.9 ± 5.6% vs. 2.4 ± 2.2%, p < 0.0001). The majority of individuals were categorized as low- and intermediate-risk, with very few men considered at high risk (Table 3).No women were classified as being at high risk.
Relationship of CCS and 10-year CHD risk
Although individuals with higher calculated 10-year CHD risk showed greater odds of higher CCS levels (Table 4),very few individuals with even moderately and severely elevated CCS were classified as high-risk by the NCEP guidelines (Fig. 1).Interestingly, one-quarter of each of the participants with CCS 100 to 399 and CCS ≥400 were classified as low risk according to the NCEP guidelines.
Eligibility for pharmacotherapy across CCS levels
The prevalence of individuals not qualifying for pharmacotherapy decreased as the CCS increased (p < 0.001). Nonetheless, 59% of the individuals with CCS ≥400 did not meet criteria for drug therapy (Fig. 2).Similarly, 309 (73%) of 426 participants with CCS ≥75th percentile would not have qualified for drug therapy. These discrepancies were particularly marked for women and young individuals compared with men and older individuals in each CCS category (Figs. 3 and 4).⇓⇓
Traditional risk factors and the calculated Framingham risk score are currently used to define the odds of incident CHD in asymptomatic individuals and to identify high-risk individuals who are candidates for preventive pharmacotherapy (4,5). The decision-making process to determine the intensity of primary prevention is based on guidelines that extrapolate conclusions from population studies; however, such assessments might fall short if relied upon too strictly for individual patient management (21–24). The use of NCEP-III guidelines alone in our asymptomatic study population would have resulted in withholding of drug treatment in 65% and 59% of the individuals with CCS ≥100 and CCS ≥400, respectively, both important markers of risk for future CHD (19).
The relative weakness of the NCEP-III guidelines to detect high-risk individuals was recently highlighted by the work of Akosah et al. (6), who showed that 75% of 222 asymptomatic young adults presenting with their first and unheralded myocardial infarction would not have been considered candidates for statin therapy in the days preceding the event (6). Furthermore, Hecht et al. (7) demonstrated that only 59% of 304 asymptomatic women submitted to EBT screening were correctly identified by NCEP guidelines as either high or low risk.
Whereas the decision to initiate drug therapy currently rests on the basis of risk factors alone, an alternative—and potentially more accurate—approach could be to target patients with extensive subclinical atherosclerosis that provides the framework for clinical manifestations (24). The final evidence that such an approach would be preferable is still missing. Nonetheless, a few moderate- to large-size studies (16,17,25) have already shown the incremental prognostic value provided by anatomical information (i.e., coronary calcium) over traditional risk factors. Receiving operator characteristic curves were used to assess the relative contribution of traditional risk factors and CCS for the prediction of myocardial infarction, cardiac death (17,25), and all-cause mortality (16) in a total of approximately 12,000 asymptomatic individuals. All studies consistently demonstrated the incremental prognostic value of CCS beyond traditional risk factors to predict outcome, establishing a potentially very important role for CCS. There is, however, an important limitation inherent in the analyses mentioned above that deserves mention. In each of the studies, the greater predictive power of CCS was demonstrable in patients at intermediate pre-test probability of disease on the basis of conventional risk factors. Because the likelihood of developing CHD increases with age and chronological age is the most weighted of the risk factors considered in the Framingham equations, young individuals are less likely to be classified at intermediate and high risk compared with older subjects. In recognition of this important limitation of the scoring algorithms, Grundy (26) suggested that “calcium scores [be] used…to replace age as a risk factor in Framingham risk equation,” because they are a reliable measure of coronary plaque burden. Although CCS screening could not be proposed for all adults, for obvious cost and safety concerns, testing might be of great additional help in refining risk assessment in young patients (as defined in this article) and women who are at least at low to intermediate risk (approximately 6% CHD risk at 10 years). This would limit the number of “missed opportunities” to aggressively treat patients at greater risk to develop events. Evidence to support the latter notion already exists. In a prior study (11), 70% of the myocardial infarctions and deaths occurred in patients with a baseline CCS >75th percentile, whereas the event rate in those with absent coronary calcium has been consistently reported to average 0.1% to 0.2% per year (11,16,17,25). Furthermore, in a recent analysis comparing cardiovascular outcome in genders after EBT screening, women seemed to benefit incrementally more than men from identification and quantification of plaque burden (27).
The results of our study extend prior reports on the limitations of the Framingham algorithms as a tool to estimate risk in the individual patient. After assessing eligibility for pharmacotherapy across a spectrum of CCS in a large population (n = 1,611) of asymptomatic men and women, we concluded that, despite the presence of a substantial atherosclerotic burden, a large number of individuals (especially young subjects and women) would not qualify for pharmacotherapy. Our contention that risk assessment might benefit from the implementation of imaging modalities for atherosclerosis, although supported by several pieces of evidence already present in the literature, awaits final demonstration from the completion of longitudinal studies such as the Multi-Ethnic Study of Atherosclerosis (28) and the Heinz-Recall Study (29), even though such studies might not be completed until the end of the current decade.
Our study presents several limitations. The authors acknowledge that the purpose of risk assessment in NCEP-III is to predict CHD and not coronary atherosclerosis; however, recent studies have provided strong support for the relationship between increasing CCS (i.e., atherosclerosis) and risk of future CHD events. The study population likely showed a selection bias, because subjects might have been strongly motivated to assess their CHD risk and could afford the expense associated with undergoing EBT screening. Also, the study population was mainly composed of Caucasians, and the findings might not apply to other racial groups.
In summary, our study highlights the limitation of NCEP-III guidelines in identifying asymptomatic individuals, especially women and the young, who harbor significant coronary atherosclerosis and are likely at high risk for CHD. For these individuals, assessment of plaque burden could help refine risk assessment and, as a result, have a significant impact in the prevention of the leading cause of death in the U.S. population.
- Abbreviations and Acronyms
- coronary calcium score
- coronary heart disease
- electron beam tomography
- high-density lipoprotein
- low-density lipoprotein
- National Cholesterol Education Program
- Received May 10, 2005.
- Revision received July 1, 2005.
- Accepted July 19, 2005.
- American College of Cardiology Foundation
- American Heart Association
- Grundy S.M.,
- Pasternak R.,
- Greenland P.
- National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)
- Grundy S.M.,
- Cleeman J.I.,
- Merz C.N.,
- et al.
- Akosah K.O.,
- Schaper A.,
- Cogbill C.,
- et al.
- Hecht H.S.,
- Superko R.
- Rumbeger J.A.,
- Simons D.B.,
- Fitzpatrick L.A.,
- Sheedy P.F.,
- Schwartz R.S.
- Simons D.B.,
- Schwartz R.S.,
- Edwards W.D.,
- Sheedy P.F.,
- Breen J.F.,
- Rumberger J.A.
- Raggi P.,
- Callister T.Q.,
- Cooil B.,
- et al.
- Arad Y.,
- Spadaro L.A.,
- Goodman K.,
- et al.
- Arad Y.,
- Spadaro L.A.,
- Goodman K.,
- Newstein D.,
- Guerci A.D.
- Kondos G.T.,
- Hoff J.A.,
- Sevrukov A.,
- et al.
- Agatston A.S.,
- Janowitz W.R.,
- Hildner F.J.,
- et al.
- Grundy S.M.
- Pashkow F.J.
- Berman D.S.,
- Wong N.D.
- Bild D.E.,
- Bluemke D.A.,
- Burke G.L.,
- et al.
- Schmermund A.,
- Mohlenkamp S.,
- Stang A.,
- et al.