It has been appreciated for many years that the fatty content of atherosclerotic plaque derives from the cholesterol carried by lipoprotein particles in the blood. The clinical importance of serum cholesterol level was established by epidemiological studies conducted during the 1950s and 1960s.^, The efficacy of diet and drugs in lowering serum cholesterol and reducing the incidence of coronary heart disease (CHD) was shown by large randomized clinical trials conducted in the 1970s and 1980s.^,, Clinical trials have also shown that pharmacological lowering of serum cholesterol can result in regression of atherosclerotic plaques. These observations, together with an extensive body of biological, observational and clinical trial data, form the basis for consensus recommendations and guidelines from the National Cholesterol Education Program (NCEP) Adult Treatment Panel (ATP) for the diagnosis and treatment of hyperlipidemia. More recent secondary^8,9,10,11 and primary^12,13 prevention trials have strengthened the evidence on which these recommendations rest and extend the demonstration of treatment benefits to new groups of patients.

The goal of treatment in the individual patient is to reduce the patient’s risk of CHD. This benefit is measured as the reduction in absolute risk of CHD (individual attributable risk). Table 1, which is based on data from the Framingham Heart Study, illustrates that atherosclerosis and CHD risk are multifactorial. Hyperlipidemia is therefore treated in the context of overall CHD risk reduction with emphasis on patient education and lifestyle change (e.g. smoking cessation and a regular exercise program). The relationship between lipid level and CHD risk is curvilinear, which means that the higher the lipid level, the greater the CHD risk reduction for a specified absolute reduction in lipid level. Finally, the benefit of lipid reduction is greatest for patients at the highest level of CHD risk (Table 1).

These principles of medical practice with individual patients may be distinguished from the principles of public health practice in populations, where most cases of CHD occur among those with mild hypercholesterolemia (or mild hypertension) and only moderately increased CHD risk. This is because serum cholesterol (like BP and overall CHD risk level) is more or less normally distributed, so that there are only a relatively few people at the high end of the distribution but many people within one standard deviation of the mean. The benefit of risk reduction in populations (population attributable risk) depends both on the absolute risk reduction and the prevalence of the risk factor in the population. Even a substantial reduction in the level of a rare risk factor will have little population impact, while a modest reduction in a common one, such as mild hypercholesterolemia, may have substantial population impact.

The NCEP ATP⁷ recommends screening all adults every five years for hypercholesterolemia by obtaining a fasting lipid profile. Fasting (12-14 hour fast except for water) is required because post-prandial triglyceridemia is not fully cleared for 8 hours in normal people and longer in people with delayed TG clearance. The laboratory measures triglycerides (TG), total cholesterol (TC), and HDL cholesterol; when TG < 400 mg/dL, LDL cholesterol is calculated using the Friedwald equation:

Several classification schemes for lipid abnormalities are used, but as a practical matter patients will have elevated LDL, elevated TG or both. Some may have concomitant low HDL. Hyperlipidemic patients should have liver and renal function tests, glucose and TSH level in order to detect treatable secondary causes. Drugs that increase LDL cholesterol and lower HDL cholesterol include anabolic steroids, coricosteroids, and progestins. The NCEP classification of patients is based on LDL cholesterol level (Table 2).

The goal of treatment is to lower the LDL level, and treatment decisions are based on LDL level plus presence or absence of risk factors as outlined in Table 3.

It is clear that risk factors vary in degree, that not all risk factors are equivalent, and that 2 risk factors and more than 2 lead to different risk. The ATPIII Guidelines recommend calculating a Framingham risk assessment for patients without known vascular disease or diabetes who have 2 or more risk factors. This can be done by inserting the patient’s age, cholesterol, HDL, and blood pressure into an equation. The program is available at http://hin.nhlbi.nih.gov/atpiii/riskcalc.htm and can be used online or downloaded to any computer or palm pilot. Three levels of risk are >20%/10 years (denoted CHD risk equivalent), 10-20% risk (same as 2+ risk factors), and low risk <10% (treated as 1 or zero risk factors) (see below).

Weight reduction may be difficult to achieve. Faced with a choice of caloric restriction or isocaloric substitution of low for high saturated fat in the diet, many patients find the latter is far easier to sustain. Successful management will focus on changes that are acceptable to patients as well as recommended by experts. The role of the patient and the role of the physician (or nutritionist) in implementing these dietary recommendations will vary widely; many people have successfully implemented them without medical assistance, while others need educational and motivational help and others will not follow them no matter what. Clinical studies have shown a biological variability in the response of serum lipid levels to dietary manipulation under closely controlled conditions; some people are nonresponders.¹⁵

Dietary advice is an integral part of medical practice. There is no brief, reliable, and valid method to assess usual diet, but it is useful to know that saturated fat is a stronger determinant of LDL level than total fat or cholesterol intake.¹⁶ People eat foods, not nutrients, and the foods that are the most important sources of saturated fat for most adults are meat, milk, cheese, butter or margarine, and ice cream; these do not vary very much by age, sex, educational level or race/ethnicity. Thus, a very targeted diet history may suffice as the basis for recommendations. The focus should be on substitution of low fat (skim [0% fat] or low fat [1% or 2% fat]) for whole milk (4% fat); monounsaturated oil (olive oil) for butter or hydrogenated (hardened) vegetable oil (margarine); vegetables, legumes and grains for meat; chicken with skin and fat removed for chicken with skin and fat; chicken and turkey for sandwich meats; and fruit for ice cream. Food and meals are a major source of pleasure for most people, and a social occasion for many. Be positive. Focus on making diet healthier and better, rather than eliminating foods people like. A surprisingly large number of meals are eaten at fast food restaurants, and many more are based on prepared foods. Try to make patients aware that these foods are often very high in saturated fat (as well as sugar and salt). Encourage them to read the labels and gain control of their diet. An important issue for many patients in making food substitutions is cost; this particularly pertains to fish, which is much more expensive per pound than meat.

Metabolic ward data indicate that after two weeks on a fat-reduced diet the serum cholesterol level is stable at the diet-responsive level. ATP recommendations are to assess response by remeasuring the serum lipid levels after six weeks. The ATP encourages referral of patients to registered dieticians or other qualified nutritionists for medical nutrition therapy, which is the term for nutritional intervention by a nutrition professional.

Drug treatment is recommended for hyperlipidemic patients who do not reach treatment targets on diet. The NCEP recommends statins, bile acid resins (e.g., cholestyramine, cholestipol), or nicotinic acid (niacin) as first line treatment. Resins and niacin have undesirable side effects, and statins are widely used. Many asymptomatic patients do not tolerate bile acid resins well because of bloating, constipation and other GI complaints. Cholesevalan (Welchol®) is a pill form of resin-like drug that may be more acceptable to some patients. Niacin is contraindicated in patients with abnormal liver function tests, history of ulcer disease or gout, and it can cause worsening of glucose intolerance or hyperuricemia. Niacin is not marketed because it is not patented, but it is inexpensive and a highly useful drug, particularly for patients who must pay for their own medicines. Niacin can cause flushing approximately 30 minutes after taking the drug, especially when beginning the drug or increasing the dose. We recommend giving aspirin 30 minutes prior to niacin, beginning with a low dose of niacin (250 mg bid) and gradually (at two to four week intervals) increasing the dose to 1 to 2 gm bid. Nicotinamide, a vitamin form of niacin, is ineffective as a lipid-lowering drug. Over the counter sSlow release preparations are hepatotoxic and are not recommended.¹⁷

A prescription-only form of slow-release niacin is sold as Niaspan®. This medication is sold in 500 and 1000 mg tablets. While it also causes flushing, appreciation of this side effect may be reduced because the single daily dose is taken at night.

HMGCoA reductase inhibitors ("statins") are highly effective and have few side effects. These drugs can cause a severe myositis, which occurs more frequently in patients who are also taking fibric acid derivatives (e.g., gemfibrozil), cyclosporin, niacin, and perhaps erythromycin. If two drugs are needed, resins can be safely combined with niacin, statins or gemfibrozil. There are presently five statins available (Table 6). Cerivastatin (Baycol) was withdrawn in 2001.

All of the statins have their greatest effect at the initiation dose. Only modest incremental reductions in LDL cholesterol are generally achieved by dose escalation to the maximum dose (~7% greater LDL reduction with each doubling of the dose); , but ccost increases greatly with dose escalation. The adverse clinical effects (fatigue, myalgia, malaise, headache, pruritis) and laboratory effects (increased aminotransferase, alkaline phosphatase, bilirubin, and creatine kinase) are more likely at higher doses.¹⁸

The role of fibric acid derivatives (e.g., gemfibrozil and fenofibrate) is currently unclear. Gemfibrozil is effective in lowering TG but is less effective than the drugs just discussed in lowering LDL. Data from one large randomized placebo-controlled study¹⁹ showed a reduction in CHD events greater than could be explained by the reduction in LDL. Speculative explanations include effects on lowering of TG, raising HDL, and reducing the oxidation state of LDL. The dose of gemfibrozil is 600 mg bid. A second fibric acid, fenofibrate, is available as a 160 mg once a day treatment. Fibric acid derivatives are contraindicated in patients with hepatic or biliary disease and generally should not be given concomitantly with statins because they greatly increase the risk of myositis with these drugs. Exceptions are made in Lipid Clinic for patients with severe hyperlipidemias.

Antioxidants are currently not recommended. Laboratory data suggest that oxidized LDL is more atherogenic than native LDL and that oxidized LDL particles are ligands for "scavenger" receptors on macrophages, are taken up by these cells, and promote their conversion to cholesterol-laden foam cells characteristic of fatty streaks.²⁰ Animal studies show that antioxidants decrease lesion progression.²⁰ Two observational studies, one in women²¹ and one in men,²² found substantial reductions in CHD events in those consuming vitamin E supplements. However, the three randomized trials of antioxidants for secondary prevention of CHD showed trends toward increased adverse outcome rates in the treatment groups.^23-25

Estrogen replacement therapy in postmenopausal women lowers LDL and raises HDL levels. Estrogen also lowers Lp(a) levels and may inhibit oxidation of LDL.²⁶ Observational studies show a reduction in CHD in women who take estrogen replacement therapy, but these studies were not designed to control for selection bias (healthier women take estrogen). Two randomized trials (the HERS Study²⁷ and the Women’s Health Initiative²⁸) found no overall difference and an increase in early CVD endpoints in the treatment group, perhaps due to pro-thrombogenic or other plaque destabilizing effects of estrogen. Two placebo-controlled trials focused on atherosclerosis progression rather than CVD events have had discrepant results. One trial in older women found no reduction in the rate of progression of coronary atherosclerosis in the estrogen-treated group.²⁹ In view of the increase in risk for cancers of the breast,³⁰ ovary,³¹ and endometrium³² associated with hormone replacement therapy and the availability of more effective alternatives with less risk, estrogen replacement therapy is not currently recommended to treat hyperlipidemia or for the primary or secondary prevention of CVD.³³

After starting drug treatment, the LDL should be measured at four to six weeks and again at three months. If treatment goals have been attained, follow-up should be at 4-month intervals to monitor lipid levels and potential side effects. In stable patients on drug treatment, monitoring of lipid levels and CK can be done every four to six months.⁷

Several wo large primary prevention trials using statin medications have been reportedrecently been reported. The West of Scotland Study (WOSCOPS) compared pravastatin 40 mg/day to placebo in 6,595 diet-treated hypercholesterolemic men (no women) aged 45-64 years with LDL 155-232 mg/dl and mean total cholesterol of 272 mg/dl at entry. At average follow-up of 4.9 years, death occurred in 3.2% vs. 4.1% (treatment vs. placebo) and major coronary events in 5.5% vs. 7.9%.¹² The WOSCOPS entry criteria were such that all of the subjects had at least two CHD risk factors as defined by the NCEP (Table 1): all were male and all were age > 45. More than 40% were smokers. Almost all had hypercholesterolemia based on total cholesterol level, and almost had entry LDL level at or above the NCEP drug treatment initiation level of 160 mg/dl for those with > two or more risk factors. Thus, while the WOSCOPS results are impressive, they are consistent with current NCEP guidelines.

The AFCAPS/TexCAPS study results showed treatment benefits in people at lower LDL level at entry. The study group also had low HDL levels. This study compared lovastatin 20-40 mg/day to placebo in 5,608 men and 997 post-menopausal women with entry total cholesterol 180-264 mg/dl (mean 221mg/dl), LDL 130-190 mg/dl (mean 150 mg/dl) and HDL < 45 mg/dl (mean 40 mg/dl) for men and < 47mg/dl (mean 36 mg/dl) for women. At average follow-up of 5.2 years, first acute major coronary events occurred in 3.5% vs. 5.5% (treatment vs. placebo), revascularization procedures in 6.2% vs. 9.3%, fatal and non-fatal MI in 3.3% vs. 5.6%, and fatal CVD events in 1.0% vs. 1.4%.¹³

A key point in interpreting the AFCAPS/TexCAPS results is that almost all of the patients had two or more CHD risk factors (e.g., 85% males, 100% > age 45 for males and > 55 for females, 12% smoked, and 22% had hypertension). The authors reported a subgroup analysis for LDL level at entry grouped by tertile. The numbers of primary endpoint events (treatment vs. placebo) were 37 vs. 54 in those with LDL < 142 mg/dl; 33 vs. 52 in those with LDL 143-156 mg/dl; and 46 vs. 77 in those with LDL > 157 mg/dl. Thus, approximately two-thirds of the subjects in this study had entry LDL cholesterol below the current NCEP drug treatment initiation level of 160 mg/dl for those without CHD but with > two CHD risk factors. These data support initiating drug treatment in men over age 45 with low HDL (< 45 mg/dl) and LDL in the 130-159 mg/dl range.

A meta-analysis of the four reported randomized primary prevention trials of lipid-lowering drugs^4,12,13,19 found a 30% reduction (95% CI, 0.62 to 0.79) in CHD event risk over five to seven years but not in all cause morbidity (0.94; 95% CI, 0.81 to 1.09).³⁴ While these randomized trials produced clinically significant reductions in total and LDL cholesterol in the treatment groups, there is speculation that the clinical benefits observed may in part be mediated by other mechanisms. These potential mechanisms include favorable effects on endothelial function,³⁵ anti-atherothrombotic effects on platelets,³⁶ and plaque stabilization.³⁶

Several areas of controversy remain. One is whether to initiate drug treatment in people without CHD and with fewer than two CHD risk factors who have LDL in the 130-160 mg/dl range. A second is how to incorporate HDL levels more effectively into treatment decisions in the absence of currently available drugs that have clinically important HDL-raising effects. Use of the Framingham risk equation places greater emphasis on HDL than previous guidelines that only used low HDL as an additional risk factor. A third is how to incorporate information on coronary calcium score obtained in subclinical CHD screening using electron beam or other special forms of CT scanning. Many experts would initiate statin treatment in patients with LDL in the 130-159 mg/dl range who have coronary calcium, as well as treating other modifiable risk factors.

Several recently completed clinical trials have helped to redefine the role of statins in patients with angina or who are who have had myocardial infarction or coronary revascularization. The 4S Study compared simvastatin (20-40 mg/day) to placebo in 4,444 diet-treated men and women with a history of angina or MI and total cholesterol 213-310 mg/dl (mean 261 mg/dl) at entry. At 4.9-6.3 years of follow-up, death occurred in 8% vs. 12% (treatment vs. placebo), major coronary events in 15.9% vs. 22.6%, and stroke in 3.4% vs. 4.6%.⁸⁸ The LIPID Study compared simvastatin (40 mg/day) to placebo in 9,014 diet-treated men and women with a history of hospitalization for unstable angina or myocardial infarction and total cholesterol 155-271 mg/dl (median 218 mg/dl) at entry. The median LDL at entry was 150 mg/dl (interquartile range 130-170) and the median HDL was 36 mg/dl (interquartile range 31-42). At mean 6.1 years of follow-up, death occurred in 14.1% vs. 11.0% (treatment vs. placebo) and death from coronary heart disease in 8.3% vs. 6.4%. Nonfatal coronary events, revascularization, and stroke were also reduced.⁹

The CARE Study compared pravastatin (40 mg/day) to placebo in 4,159 diet-treated men and women with a history of MI. Lipid entry criteria were total cholesterol < 240 mg/dl and LDL 115-174 mg/dl (3.0-4.5 mmol/L). Mean LDL cholesterol at baseline was 139 mg/dl, and treatment lowered LDL to a mean level of 98 mg/dl. At mean 5 years of follow-up, fatal or non-fatal MI occurred in 10.2% vs. 13.2% (treatment vs. placebo) and stroke in 2.6% vs. 3.8%.¹⁰ A subgroup analysis of the 1,283 patients aged 65-75 years in the CARE study reported major coronary events in 19.7% vs. 28.1% (treatment vs. placebo), coronary death in 5.8% vs. 10.3%, and stroke in 4.5% vs. 7.3%.¹¹

Taken together, these data strongly support the NCEP ATP recommendation that patients with CHD should have drug treatment initiated if, on diet, the LDL cholesterol is > 130 mg/dl, with a treatment target of < 100 mg/dl. Subgroup analysis in the CARE study showed that event rates decreased progressively as LDL cholesterol levels fell from 174 to 125 mg/dl, but from 125 to 71 mg/dl, CHD events did not decline further.³⁷ Thus, one interpretation of the CARE study is that the published data do not warrant escalation of drug treatment to achieve treatment goals beyond these recommendations. We note, however, that some experts believe that the CARE Study was underpowered to show a difference in this subgroup and would initiate treatment in all MI patients with LDL of > 120 mg/dl, or even lower, and would aim to achieve an LDL level < 100 mg/dl. A more recent study (Heart Protection Study) has been presented in abstract but not yet published in full (AHA 2001, Late Breaking Clinical Trials). This study included ~20,000 patients in the UK with risk factors or CHD; half of these patients were treated with a statin (simvastatin 40 mg). All groups including women, the elderly, and patients with LDL < 125 obtained benefit from cholesterol-lowering therapy.

Hypertriglyceridemia as an isolated disorder appears from clinical studies to cause little or no increase in CHD risk,³⁸ but there is controversy on this point because VLDL particles are catabolized to LDL, so that increased VLDL is often associated with increased LDL cholesterol. Hypertriglyceridemic patients also often have low HDL levels. Renal failure, diabetes mellitus, alcohol, estrogens and severe obesity are associated with hypertriglyceridemia. TG levels of > 1000 mg/dl are associated with pancreatitis. Drug treatment is indicated in such patients or in patients with lesser TG elevations but a history of pancreatitis.⁷ Niacin, gemfibrozil, and fish oil (1000 mg, 1-2 tid with meals) are acceptable alternatives. Weight reduction, reduction in alcohol intake and increased physical activity should be stressed.

Low HDL (< 40 35 mg/dl) appears from epidemiological data to be an independent risk factor for CVD,³⁹ although one report described a small number of patients with very low HDL and no CVD.⁴⁰ Treatment of low HDL in the presence of normal LDL levels is not recommended, but low HDL is an additional risk factor in profiling patients with high LDL. Factors that increase the level of HDL include female gender, exogenous estrogens, regular exercise and moderate alcohol intake. Male gender and cigarette smoking decrease HDL. We do not recommend that patients drink alcohol, because of the many deleterious effects, but we also do not discourage moderate alcohol intake. Exercise should be encouraged. The VA High-Density Lipoprotein Intervention Trial (VA-HIT) randomized 2,531 men with known CHD and low HDL (mean 32 mg/dl) and low LDL (mean 111 mg/dl) to gemfibrozil 1,200 mg/day vs. placebo. Mean HDL levels were 33.4 mg/dl in the gemfibrozil group vs. 31.7 mg/dl in the placebo group. TG levels were 101 mg/dl vs. 156 mg/dl, respectively. LDL levels did not differ (113 mg/dl in both groups). CHD events were reduced by 11% for each 5 mg/dl increase in HDL (P=0.02).⁴¹ The discrepancy between the modest HDL-raising effect and the more substantial treatment effect on CHD risk has been interpreted as evidence for additional mechanisms of action. This study gives some support to use of gemfibrozil in patients with CHD who have low HDL and low LDL. In patients with low HDL and high LDL, statin treatment is the preferred strategy.

Dyslipidemia in diabetes is a complex disorder that typically comprises high LDL, low HDL and high TG. Total cholesterol may be elevated or may be normal, despite the presence of lipid abnormalities. All diabetics should have a fasting lipid profile. Peripheral insulin resistance in type 2 diabetes produces hyperinsulinemia, which in turn stimulates hepatic production of TG-rich lipoproteins. The currently recommended diet for diabetics is consistent with behavioral management of hyperlipidemia, in that it calls for limiting total fat intake to < 30% and saturated fat intake to < 10% of energy intake and dietary cholesterol to < 300 mg/day.⁴² Weight reduction, which would be useful for most diabetics and most hyperlipidemic patients, and which is recommended as part of the behavioral management of both conditions, can be difficult to achieve by caloric reduction alone. Exercise, which raises HDL independent of effects on weight loss, should be encouraged. Niacin can worsen glucose intolerance and should generally be avoided in patients with diabetes. Otherwise, drug treatment may be as for nondiabetics.

The metabolic syndrome, also termed insulin resistance syndrome, Syndrome X, or Reaven’s Syndrome, refers to patients with truncal obesity, elevated fasting insulin, peripheral insulin resistance, elevated BP and dyslipidemia characterized by low HDL, high TG and sometimes high LDL. This syndrome has been described in children as well as adults, in the absence of frank diabetes. Management is primarily behavioral (exercise and weight loss), as for dyslipidemia in diabetes. The metabolic syndrome has increasingly been recognized as a consequence of obesity and a central issue in management of CHD risk. The ATP III Guidelines give special emphasis to the metabolic syndrome in the pathogenesis of dyslipidemia and as a target of the TLC strategy of behavioral management. ATPIII has a secondary goal of therapy, reduction of non-HDL cholesterol, i.e. LDL + VLDL, to a level that is 30 mg/dl greater than the goals for LDL management. This suggests a more vigorous approach to treatment of hypertriglyceridemia.

Older patients, specifically those over age 65, have more CHD events, thus increasing the potential benefit (reduction in absolute risk in an individual), but epidemiologic data indicate that TC levels become less strongly predictive of CHD as age increases. Low HDL may be a more important CHD risk factor than high LDL in older patients.⁴³ The NCEP ATP guidelines are intended to apply to all adults over age 20. A study in patients over 65 years of age (mean age 71) found that diet plus lovastatin 20 mg/day vs. 40 mg/day was equally effective in lowering LDL cholesterol (24% vs. 28%).⁴⁴ Subgroup analysis of the 4S trial⁸ showed that statin treatment in patients over 65 leads to a similar reduction in second CHD events as was observed in younger patients. No recommendations have been published modifying the NCEP guidelines for the elderly, whether healthy or with other comorbid conditions. Clinical judgment regarding treatment for hyperlipidemia should be similar to mild hypertension in that both are asymptomatic conditions that elevate CHD risk, while treatment is costly, has side effects, and must be monitored.

Younger patients, specifically men under the age of 35 and premenopausal women without other risk factors whose LDL is 190-220 mg/dl, should be managed with diet and other life-style modification, but drug treatment may be delayed.⁷ As with older patients, there are no clinical trial data directly addressing treatment benefits in this group.

Familial hypercholesterolemia (FH) is an autosomal dominant genetic disorder that affects the function of the LDL receptor.⁴⁵ Homozygotes (approximately 1 in 1 million persons) have LDL levels that often exceed 500 mg/dl, develop planar cutaneous xanthoma in childhood, fulminant CAD, often in childhood, and aortic stenosis from valvular cholesterol deposition. Heterozygotes (approximately 1 in 500 persons) have two to three-fold elevations of LDL, often have strong family histories of early onset CAD, tendon xanthomas (diffuse or nodular thickening of the Achilles’ tendons), xanthelasmas, and arcus corneae. Tendon xanthomas are virtually diagnostic for FH, while the latter two findings are nonspecific. Heterozygous FH accounts for about 5 percent of patients under age 60 with MI. FH describes a set of different mutations in the LDL gene that characterize different families and groups, explaining in part the variable severity of disease and age of onset. Treatment often requires a combination of strict diet, careful attention to other risk factors, and two or more drugs. Ileal bypass, plasmapheresis, and liver transplantations have been used experimentally in homozygous FH patients. , and tThe condition may be a target for gene therapy and, since it is a genetic disease due to a well-defined single mutation, it is amenable to prenatal diagnosis.