Lipid-Lowering Therapy: What the New NCEP

From Center For Clinical Age Management, Inc.

Cardiac Risk Factor
Lipid-Lowering Therapy: What the New NCEP
By By Benjamin J. Ansell, MD, FACP
Sep 6, 2002, 7:41am

Cardiovascular risk factor modification is typically not considered an urgent matter, and certainly not an emergency. However, the devastating effects of coronary heart disease (CHD) and stroke have made it a top public health priority. Recent evidence suggests that lipid-lowering therapy offers perhaps the greatest potential for coronary risk reduction among current treatment regimens, that there are both short- and long-term cardiovascular benefits to such therapy, and that most at-risk patients are not presently receiving appropriate treatment. This justifies broad-based efforts to educate health care providers, including cardiovascular specialists, generalists, emergency physicians, physician assistants, and nurse practitioners, about the appropriate role for lipid-lowering therapy.

The new National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP-III) guidelines reflect the most recent effort to accomplish that objective. These guidelines, which will be the focus of this article, are based on the results of numerous studies since ATP-II that revealed the benefits of lipid-lowering therapy in patients at increased risk for cardiovascular events.

HIGHLIGHTS OF THE NEW GUIDELINES

The implications of the new guidelines for patients with the most severe forms of dyslipidemia are obvious. The real challenge of evidence-based practice, however, is to decide when to use pharmacotherapy in patients with borderline levels of risk. Fortunately, the new NCEP recommendations provide much more direction in this area than previous guidelines. They also include more specific recommendations with respect to lifestyle changes for all patients, in terms of both diet and exercise, as well as a renewed emphasis on aggressively lowering low-density lipoprotein (LDL) cholesterol levels below 100 mg/dl in patients with CHD and other high-risk patients.

ATP-III also addresses the role of lipid-lowering therapy in primary prevention—that is, in at-risk individuals who have not yet been diagnosed with cardiovascular disease but who can be treated to reduce their risk of heart attack and stroke. In such patients, the number and severity of their CHD risk factors suggests the appropriate LDL treatment target levels. Some novel markers and tests may also prove helpful in trying to identify patients with an ambiguous risk profile who should be treated, but the precise role of such testing needs further clarification.

Another lipid treatment goal in the new NCEP guidelines is called the "non-HDL" level, which is determined by subtracting the high-density lipoprotein (HDL) cholesterol level from the total cholesterol level. It was chosen by the ATP-III as a target because it includes all of the atherogenic lipoproteins, including LDL, intermediate-density lipoprotein (IDL) cholesterol, very-low-density (VLD) lipoprotein cholesterol (mostly triglycerides), and remnant particles. Non-HDL is a secondary target for intervention in patients who are at or below their LDL target but have hypertriglyceridemia. The non-HDL goal for an individual is 30 mg/dL higher than his or her LDL target.

The new guidelines place particular emphasis on populations that have traditionally been underrepresented in both guidelines and clinical trials in the United States. These include women, who suffer as many myocardial infarctions as men do each year; diabetic patients, who typically have the same risk of heart disease as individuals who have had a myocardial infarction; and the elderly, who account for 85% of all cases of cardiovascular disease.

Lastly, the guidelines recognize that many patients have complex dyslipidemic profiles that often require combination therapies.

STILL VALID ATP-II RECOMMENDATIONS

Many recommendations from ATP-II have since been validated by recent clinical trials and have been incorporated into the ATP-III guidelines. One such recommendation is making reduction of LDL cholesterol levels the primary target of therapy, with the degree of reduction based on the underlying CHD risk of a patient. Those who have known CHD or atherosclerosis should undergo the most aggressive treatment; those with multiple risk factors require less aggressive treatment; and those who have only one risk factor, or none, in addition to dyslipidemia warrant the least aggressive treatment. All patients, according to ATP-III, should be advised to achieve ideal body weight and to exercise at least three or four times a week for 20 to 30 minutes at a time. This is referred to in the new guidelines as "therapeutic lifestyle changes" (or "TLC").

The most dramatic changes in ATP-III relate to treatment of patients with multiple risk factors. Patients who have diabetes, for example, are not considered to simply have a risk factor for CHD in the new guidelines; they are now considered to have a risk equivalent to patients with established CHD, requiring that comparable treatment goals be set. For patients who have two or more risk factors, ATP-III suggests calculating a risk "score" based on the presence and severity of CHD risk factors to determine how intensively LDL should be reduced.

ATP-III suggests an LDL level of less than 100 mg/dl as optimal for all patients, but establishes it as a formal goal only for high-risk patients. An HDL level of less than 40 mg/dl is now considered a major coronary risk factor, up from the previous level of 35 mg/dl or less. This change will substantially increase the number of candidates for drug therapy under the new guidelines. Triglyceride goals are encompassed in the new target non-HDL level. However, reducing triglycerides to levels below 150 mg/dl will nearly always meet the non-HDL goal.

PATIENTS WITH MULTIPLE RISK FACTORS

The new NCEP guidelines set target goals that are more aggressive than preceding recommendations, reflecting recent evidence showing the benefits of such treatment in these patients. An example of such a patient is Mike, a 55-year-old dentist of Filipino descent. As is the case with many patients, Mike does not appreciate his increased level of risk. His medical history is remarkable for a cholecystectomy, and his medications include omeprazole, aspirin, pravastatin, and over-the-counter niacin. He completely abstains from tobacco and alcohol and denies any symptoms or history that suggest established CHD.

A closer look at Mike's history, though, reveals some cause for concern. A few months prior to seeing me for the first time, he was on a business trip in Chicago, where he had an episode of atypical chest pain. He saw a cardiologist there and underwent a thallium stress test, which was negative. Nevertheless, the cardiologist recognized that Mike was at increased risk for CHD because of his elevated cholesterol level. He gave him a prescription for atorvastatin, recommended that he take aspirin, and suggested that he take metoprolol for his high blood pressure, which was also noted at the time.

Mike largely ignored these recommendations because he remained asymptomatic and did not think of himself as being at immediate risk. He did not fill his prescription for atorvastatin because it was not on his insurance plan's formulary. He did not take the metoprolol either because he was concerned about the risk of erectile dysfunction. He spoke with some of his friends, including another dentist, who suggested that he begin taking over-the-counter niacin. Mike followed this suggestion and started taking 50 mg of niacin daily. When he did finally speak with a physician friend, he started taking a statin, too, that was on his plan's formulary, which was pravastatin at 20 mg daily.

When I saw Mike, his blood pressure was 130/78 mm Hg, even though he was not taking the metoprolol. He was clearly overweight, however, with a waistline of more than 40 inches and a body mass index (BMI) of 29, but his physical examination was otherwise normal except for the cholecystectomy scar. His lipid profile was clearly a concern, even with the pravastatin (see table below). His total cholesterol level was borderline-high at 226 mg/dl, his triglyceride level was very high at 309 mg/dl, and his HDL level was low at 29 mg/dl. His LDL level was 149 mg/dl, which may seem normal, but in patients with an HDL below 40 mg/dl or a triglyceride level above 150 mg/dl, LDL particles are smaller, denser, and more atherogenic. Liver function testing revealed an aspartate aminotransferase level at the upper limits of normal and an above-normal alanine aminotransferase level, while his fasting glucose was slightly elevated at 114 mg/dl.

Mike's fasting lipid and basic chemistry profile on pravastatin 20 mg daily, atorvastatin 10 mg daily, and the combination of atorvastatin 10 mg daily with gemfibrozil 600 mg daily. (All values in mg/dl, except AST and ALT, which are expressed in U/L.)

CRITERIA FOR METABOLIC SYNDROME

With patients like Mike, laboratory test results often appear relatively normal or only slightly abnormal. Clearly, though, Mike has multiple abnormalities here and is at increased risk for CHD. This is typical of patients with the so-called metabolic syndrome. The new guidelines define this syndrome as the presence of three or more of the following: abdominal obesity, a triglyceride level above 150 mg/dl, an HDL level below 40 mg/dl in men or below 50 mg/dl in women, systolic blood pressure over 130 mm Hg, and a fasting glucose level above 110 mg/dl. Mike meets all four criteria.

Mike's current statin regimen does not appear to be adequate, considering his elevated triglycerides and LDL level. While large studies involving adding triglyceride-lowering medication to statin regimens have not been completed, there is enough current evidence available to support a change in therapy. A number of options are available, including lifestyle modifications, increasing his pravastatin dose, switching him to a statin that will do a better job of lowering his LDL and triglycerides levels, adding a fibrate to reduce his triglyceride level, increasing his niacin from a relatively homeopathic dose to a more pharmacologically relevant dose, and/or considering an insulin-sensitizing agent.

Generally, a single intervention is not adequate for patients with the metabolic syndrome. Patients like Mike require a multidisciplinary approach involving his physician, a physician assistant or nurse practitioner, if available, a nutritionist, and the patient himself. I recommended that Mike see our nutritionist, start an exercise program, and switch to atorvastatin 10 mg daily for better control of triglycerides. Further evaluation of his liver function test abnormalities revealed negative viral hepatitis serologies, and an abdominal ultrasound confirmed the presence of a fatty liver. His glycosylated hemoglobin level was at the upper limit of normal at 5.9%.

ATP-III suggests that a major focus of treatment for most patients with dyslipidemia should be dietary modification. Unfortunately, vague recommendations regarding diet often produce only minimal effects; most patients assume that the goal is simply to lose weight and to eat more healthful foods. However, a recent meta-analysis of studies comparing American Heart Association (AHA) step I and step II diets with no dietary recommendation did not reveal a difference in coronary, non-coronary, or total mortality. That should not prompt a nihilistic approach to dietary therapy, but rather the use of dietary interventions that are easy to remember and more specific than just "eat a low-fat diet."

BENEFITS OF THE MEDITERRANEAN DIET

One dietary approach that has proved successful in reducing heart disease is the so-called Mediterranean diet. In the Lyon Heart Study, patients on a Mediterranean diet, compared to those on a conventional Western diet, enjoyed a significant reduction in the risk of coronary events, a composite cardiovascular endpoint, and in mortality despite the fact that a higher percentage of Western-diet patients were taking lipid-lowering medication. Substituting monounsaturated fat sources (olive, peanut, and canola oils, for example) for saturated fat sources, increasing consumption of vegetables, and eating fish regularly are all part of a Mediterranean diet plan that should benefit patients.

The new diet in the ATP-III guidelines strongly advises consumption of monounsaturated fats (up to 20% of recommended total calories come from these sources) rather than saturated fats, particularly those containing trans fatty acids. Other recommendations include increasing intake of soluble fiber, increasing consumption of fruits and vegetables, and substituting plant stanol-derived spreads for butter or margarine.

After Mike adopted the lifestyle changes I recommended and his statin dose was increased, his LDL level dropped from 149 to 95 mg/dl, his HDL level rose slightly from 29 to 32 mg/dl, and his triglycerides fell from over 300 to 240 mg/dl. With persistently abnormal HDL and triglyceride levels, we needed to address his CHD risk further. To this end, fenofibrate was added to his atorvastatin, based on the results of the Veterans Administration HDL Intervention Trial (VA-HIT), which showed that in CHD patients with LDL levels already near the goal of 100 mg/dl, there was a 22% reduction in coronary events with gemfibrozil versus placebo.

On the combination of atorvastatin 10 mg and fenofibrate 160 mg daily, Mike's HDL level improved to 35 mg/dl and his triglycerides fell to 159 mg/dl. His LDL, though, interestingly, rose to 110 mg/dl. The reason for this is that his pre-treatment LDL cholesterol was most likely small and dense, but adding fenofibrate converted it into a larger, more buoyant, and less atherogenic form. Although the LDL level often rises under these circumstance, reflecting an increase in the size of the LDL particles, the total number of particles usually does not increase.

Given the overall improvement in his lipid profile, most authorities would agree that the atorvastatin/fenofibrate combination is justified in Mike's case, even with his somewhat higher risk for myopathy or hepatitis. Fortunately, as in many patients with fatty liver, despite the concern about liver function abnormalities with combination therapy, Mike's transaminase levels paradoxically normalized as his dyslipidemia and weight status improved.

ADDITIONAL THERAPEUTIC OPTIONS

For Mike and other patients with metabolic syndrome, there are additional therapeutic options worth considering. One inexpensive, over-the-counter therapy that can reduce triglycerides is omega-3 fatty acids (fish oils), dosed from 1 to 6 grams daily, which have been shown to reduce vascular events in patients with vascular disease. Fish oils typically raise HDL and favorably impact some clotting parameters in the blood.

Another consideration would be angiotensin-converting-enzyme (ACE) inhibitors, which, based on the results of the Heart Outcomes Prevention Evaluation (HOPE) trial, might prevent the onset of diabetes in some individuals. There may also be a role for insulin sensitizers in these patients, but there are no endpoint studies to justify this as yet. Niacin would likely improve Mike's dyslipidemia, but it would require a significantly higher dosage than he had been taking and would carry a slight risk of worsened glycemic control and an increased risk of myopathy.

The new NCEP recommendations target triglycerides in patients such as Mike who exhibit hypertriglyceridemia. Increasing triglyceride levels are predictive of CHD risk, particularly in women (see graph on below).

Source: Castelli WP: Cholesterol and lipids in the risk of coronary artery disease—the Framingham Heart Study. Can J Cardiol 4 Suppl A:5A, 1998.
--------------------------------------------------------------------------------

Even within the normal range of 50 to 150 mg/dl, there is a two-fold gradient of risk with increasing triglyceride levels in women; in the 250 to 300 mg/dl range, the risk increases three- to four-fold. Associated abnormalities with hypertriglyceridemia often include insulin resistance, low HDL cholesterol, diabetes, and obesity and other features (for example, hypercoagulability and elevated serum markers of inflammation) seen in the metabolic syndrome. For this reason, once LDL target levels are achieved, triglycerides are a secondary lipid target in ATP-III as part of non-HDL cholesterol. Non-HDL targets are 30 mg/dL higher than corresponding LDL goals for all patients (see table below).

Serum Cholesterol Targets in ATP-III
--------------------------------------------------------------------------------

Risk Category LDL Goal
(mg/dl) Non-HDL Goal
(mg/dl)
CHD and CHD risk equivalent
(10-year risk for CHD >20%) <100 <130
Two or more risk factors
and 10-year risk < 20%
<130 <160
No more than one risk factor
(or none) <160 <190
Source: NCEP 2001

RELATIVELY SUBTLE RISK

Gabriella is another patient whose cardiovascular risk was relatively subtle. Gabriella, a successful Beverly Hills real estate agent, presented for evaluation of a thyroid nodule noted by a friend. Her family history suggested an increased risk for CHD because her father had a myocardial infarction at age 67. She was taking multivitamins, vitamin E, and St. John's wort. She rarely exercised but did not smoke.

Physical examination revealed only the thyroid nodule, which later proved benign. Laboratory testing for thyroid function was normal. Her LDL level was slightly elevated at 171 mg/dl, but her HDL was elevated at 88 mg/dl-high enough, in fact, that the ATP-III guidelines consider it a negative risk factor, offsetting other risk factors like hypertension or having a family history of premature CHD. Her triglyceride level was within the normal range at 159 mg/dL.

Despite the above-average LDL level, most patients with a lipid profile like Gabriella's are at decreased risk for vascular disease because of their healthy total cholesterol/HDL ratio, which, in Gabriella's case, is just under three. (Her total cholesterol is 263.) Nevertheless, a few patients with this profile do not appear to be protected from development of atherosclerosis, and Gabriella proved to be one of them. Perhaps due to LDL that is more susceptible to oxidation or defective HDL that is unable to prevent this oxidation, there is a small minority of patients with above-normal levels of HDL who are still at high risk for vascular disease.

To identify patients without traditional risk factors who are nonetheless predisposed to CHD, some clinicians have proposed measuring serum levels of highly sensitive C-reactive protein or the extent of coronary calcium deposition by electron-beam computerized tomography (EBCT). Gabriella did not wait to discuss these options with a physician; she followed a friend's advice and underwent EBCT scanning of her coronary arteries on her own. Surprisingly, the test revealed a very significantly elevated calcium score for her age—a score of 63, which for a 50-year-old female is in the 99th percentile.

Demonstration of detectable, abnormal levels of asymptomatic atherosclerosis has become a significant issue now with thoracic CT imaging so commonplace. The NCEP position here, while not encouraging the use of EBCT scanning, does endorse aggressive treatment (to LDL targets of 100 mg/dl or lower) in patients with clinically suspected atherosclerosis. In Gabriella's case, with a baseline LDL level of 179 mg/dl, this will require about a 40% reduction in LDL. Daily aspirin therapy would also be prudent, if not contraindicated. ACE inhibitor therapy might also be considered, given the cardiovascular benefits in patients at high risk for CHD without hypertension in the HOPE study.

It is important to recognize that all patients fall on a continuum of CHD risk, and that patients such as Gabriella who do not have manifest heart disease may still benefit from risk factor modification. The NCEP has suggested including as high-risk patients who qualify for the most aggressive LDL reduction those patients whose CHD risk is comparable to the risk in patients who have previously suffered a myocardial infarction. So for those who have had strokes, transient ischemic attacks, peripheral artery disease, type 2 diabetes, or a risk of having a myocardial infarction in the next 10 years that exceeds 20%, the target value for LDL is less than 100 mg/dl.

SIGNIFICANCE OF LOW HDL

Our third case presentation is Carl, a 55-year-old movie archivist at a Hollywood studio, who sought treatment for mild hypertension. At the time of his first visit to his physician, Carl was drinking three glasses of wine and smoking half a pack of cigarettes most days. He had no personal or family history of vascular disease or diabetes mellitus. His examination was unremarkable except for his blood pressure of 134/96 mm Hg.

Carl's total cholesterol level was 212 mg/dl; his HDL, 37 mg/dl; LDL, 135 mg/dl; triglycerides, 138 mg/dl; and glucose, 92 mg/dl. Until recently, in most laboratory reports, none of these laboratory values would have been considered as outside the normal range. But Carl's low HDL level is now recognized as a major risk factor, which, along with his hypertension, smoking, and age over 45 years, calls for lipid-lowering therapy. In addition, his alcohol consumption might be increasing his triglyceride levels and blood pressure. Based on the constellation of his smoking, hypertension, low HDL, and age, NCEP guidelines suggest reducing his LDL to less than 130 mg/dl.

Low levels of HDL like Carl's are actually a better predictor of coronary events than high LDL levels. Clinicians focus appropriately on lowering LDL to reduce coronary risk based on the benefits seen in clinical trials, but even when LDL is below 100 mg/dl, there is still an undue degree of risk associated with a low HDL (see graph below). This supports the use of therapy in these patients to both lower LDL and raise HDL.

Adapted from: Castelli WP, et al.: Incidence of coronary heart disease and lipoprotein cholesterol levels. The Framingham Study. JAMA 256:2835, 1986.

To date, the best strategy for reducing CHD risk in patients with low HDL levels is to reduce their LDL levels, based on the Air Force/Texas Coronary Atherosclerosis Prevention Study (AF/TexCAPS). This was a randomized trial of lovastatin versus placebo in middle-aged men and women with mean HDL levels of 38 mg/dl at study entry. In this trial, lovastatin treatment was associated with a reduction in LDL levels from 150 to 115 mg/dl, along with an increase in HDL from 38 to just under 40 mg/dl. This produced a 37% relative risk reduction in coronary events and a 40% reduction in myocardial infarction. Many of these low-HDL patients, therefore, actually had latent but treatable cardiovascular disease.

Carl was started on a statin and an ACE inhibitor and was given recommendations for dietary changes and an exercise program. His lipid profile improved significantly over the next several months; both his total cholesterol and LDL dropped to more acceptable levels (159 and 91 mg/dl, respectively) and his HDL showed a modest increase (to 39 mg/dl). Although the new NCEP guidelines do not suggest a specific HDL target, it was still encouraging to see it increase with treatment.

In the AF/TexCAPS study, the patients who benefited most significantly were those whose baseline HDL levels were below 40 mg/dl. These patients enjoyed a relative risk reduction in coronary events of 45% over the five-year study period. Patients with HDL levels over 40 mg/dL had relatively little risk reduction (15%). The study also showed a significant decrease in CHD risk in placebo patients whose HDL levels were more than 40 mg/dL versus those below 40 mg/dl. This was a major rationale for broadening the definition of low HDL as a risk factor in the ATP-III guidelines from 35 to 40 mg/dl.

CONCEPT OF GLOBAL RISK

Of course, HDL is just one determinant of CHD risk. It is important to remember that the most accurate method for predicting an individual's CHD risk is to assess a variety of risk factors. The NCEP suggests that this is a prudent strategy for patients with two or more CHD risk factors, using what is termed a "global risk assessment" score, based on the Framingham Heart Study (see tables below). In calculating an individual's risk of CHD in the next 10 years, family history is not considered because it did not prove to be an independent risk factor in the Framingham study. Patients with diabetes are all considered high risk, so a global risk score is not calculated. However, assigning a scale of points for total cholesterol, HDL, blood pressure, smoking history, and age allows the clinician to predict risk in other patients, based on the total point score.

This risk estimation can prove useful in patient counseling, allowing the clinician to suggest to a 45-year-old patient that his risk is comparable to the average 65-year-old, for example. In Carl's case, he scores eight points for his age of 55, three points for his total cholesterol level on presentation of 212 mg/dl, two points for his initial HDL level of 37 mg/dl, one point for his blood pressure of 134/96, and three points for smoking. That comes to a total score of 17 points, which corresponds to more than a 30% risk of CHD in the next ten years. This could be a powerful incentive for risk factor modification. (A woman with identical risk factors would score 20 points, corresponding to an 11% 10-year risk of CHD.)

Global risk scores can be used to assign patients with multiple risk factors to risk categories, based on these new NCEP recommendations, which will then help guide treatment. Patients with a 20% or greater 10-year risk of CHD should be treated to achieve an LDL target goal below 100 mg/dL. Patients with an intermediate (10% to 20%) 10-year risk should have an LDL target below 130 mg/dl. For those who have one or fewer risk factors, which carries less than a 10% risk of CHD over 10 years, the LDL target is less than 160 mg/dl.

While it is advisable to treat low- and moderate-risk dyslipidemic patients with lifestyle changes initially, most high-risk patients should begin drug therapy at the same time. This is especially true for those patients with symptomatic CHD. Waiting months to begin lipid-lowering therapy exposes these patients to excessive and unjustifiable CHD risk.

In the recent Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (MIRACL) trial, initiating high-dose atorvastatin versus placebo at the time of hospitalization for an acute coronary syndrome proved to be beneficial in reducing ischemic events after just 16 weeks. The 50% relative risk reduction for stroke that was found in this study is the greatest risk reduction demonstrated to date in statin trials. Surprisingly, these results were seen despite a relatively low mean LDL enrollment level of 124 mg/dl among these patients. The results of the MIRACL trial suggest that there is, in fact, an urgency to starting lipid-lowering therapy in high-risk patients.

EXPANDED AWARENESS

The role of lipid-lowering therapy has expanded as our awareness of its benefits has grown in recent years. We are fortunate to have a new set of NCEP guidelines to direct therapy and to identify certain patients, such as those with diabetes and those with significantly elevated risk profiles, whose LDL target goal should be less than 100 mg/dl. Clinicians need to weigh the severity of various risk factors in determining a patient's risk. While the new guidelines downplay the role of family history as an independent risk factor in global risk calculation, there is strong emphasis on reducing LDL and non-HDL cholesterol to reduce CHD risk. By utilizing these guidelines, clinicians will now be better able to identify and treat the estimated 65 million Americans who are candidates for some form of lipid-lowering treatment.

SUGGESTED READING

Ansell BJ: Developing a clinical strategy for cholesterol management in an era of unanswered questions. Am J Cardiol 88:25F, 2001.

Castelli WP: Lipids, risk factors, and ischaemic heart disease. Atherosclerosis 124:S1, 1996.

De Lorgeril M, et al.: Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation 99:779, 1999.

Downs JR, et al.: Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/ TexCAPS. Air Force/Texas Coronary Atherosclerosis Prevention Study. JAMA 279:1615, 1998.

Executive summary of the 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). JAMA 285:2486, 2001.

Rubins HB, et al.: Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. N Engl J Med 341:410, 1999.

Schwartz GG, et al.: Effects of atorvastatin on early recurrent ischemic events in acute coronary syndromes: the MIRACL study: a randomized controlled trial. JAMA 285:1711, 2001.

Wilson PW, et al.: Prediction of coronary heart disease using risk factor categories. Circulation 97:1837, 1998.