University of South Alabama (R.A.K.), Mobile, Alabama 36688; and University of Alabama (A.O.), School of Medicine, Birmingham, Alabama 35205
Address all correspondence and requests for reprints to: Robert A. Kreisberg, M.D., University of South Alabama, 307 North University Boulevard, CSAB 170, Mobile, Alabama 36688-0002. E-mail: rkreisberg{at}usouthal.edu.
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
There is a much better understanding of the molecular changes in vascular wall biology that are produced by the traditional CHD risk factors that initiate, sustain, and amplify the atherosclerotic process. Genetic factors are important, not only because they predispose to the traditional and emerging risk factors expressed by patients [e.g. hypertension, diabetes, low high-density lipoprotein-cholesterol (HDL-C), increased low-density lipoprotein-cholesterol (LDL-C), etc.], but also because they probably influence the intrinsic response of the vessel wall to external forces and factors that initiate and perpetuate the atherosclerotic process. Those exceptional persons who have multiple CHD risk factors but no disease may have a low-intensity inflammatory response and are therefore resistant, whereas those with few or no obvious risk factors, but advanced CHD, may have an excessive inflammatory response.
Clinical trials have unequivocally demonstrated that treatment of dyslipidemias reduces cardiovascular (CV) events (7). This protection derives from a variety of mechanisms, culminating in plaque stability (5). The lipid hypothesis of atherosclerosis originally related to total and LDL-C but has become much more complicated because of the atherogenicity of other apolipoprotein ß-containing lipoproteins [very LDL (VLDL), VLDL and chylomicron remnants, intermediate-density lipoprotein, and lipoprotein (a)] and low HDL-C. Qualitative differences in LDL and HDL that relate to size and buoyancy may also influence the atherosclerotic process. There is a continuum of interest among physicians in detecting and treating lipid/lipoprotein abnormalities. At one extreme are those who do not identify patients with lipid abnormalities that predispose to CHD and who, when they do identify patients, do a poor job of treating them. At the other extreme are physicians who are far out in front of the data looking for and aggressively treating subtle qualitative lipoprotein differences/abnormalities before there is objective data that it makes any difference. It is within this framework that this article will be read; for some, it will be far too complicated and for others, much too simplistic. Our intention is to provide information that is useful to effectively manage dyslipidemia on the basis of the best current evidence.
The National Cholesterol Education Program (NCEP) has provided guidelines for identification and treatment of cholesterol and other related disorders since 1988. For many of the investigators and physicians who work in the field, the recommendations of the Adult Treatment Panels (ATP) have been too conservative, but they have always tried to base their guidelines on objective data. To paraphrase, there are things in medicine that make perfect sense but are wrong (8), witness the recent results of the Womens Health Initiative (9).
The ATP III published its last report in 2001 and reaffirmed the importance of the established CHD risk factors (10). In addition, it stratified LDL-C and triglyceride levels (Table 1), recommended use of the non-HDL-C to guide therapy for patients with TG of at least 200 mg/dl after LDL-C levels were optimized, prophetically emphasized the importance of the metabolic syndrome, proposed use of the Framingham Tables for predicting CHD events for patients with at least two CHD risk factors, and identified CHD equivalents as surrogates for CHD. Those with CHD or its equivalents are generally at high risk, i.e. at least 20% probability of a coronary event in 10 yr. Those with no more than one risk factor are at low risk, i.e. less than 10% in 10 yr. For those with at least two risk factors, Framingham tables are recommended to estimate 10-yr risk (10). The report also listed a number of potentially important emerging lipid and nonlipid risk factors (Table 2
), some of which will be discussed in this review. There is considerable observational data relating some of the emerging risk factors to CHD, but for others, there is little data. There are virtually no randomized controlled trial (RCT) data on therapies directed at modification of these potentially important factors, and consequently specific recommendations are neither feasible nor, perhaps, desirable. Fortunately, many of these risk factors are amenable to therapeutic lifestyle changes and/or are modified by available hypolipidemic agents. The presence of these risk factors can be used by physicians to intensify the use of diet, exercise, weight loss, and pharmacological agents as monotherapy or combination therapy to achieve optimum control of LDL-C, HDL-C, triglyceride, and non-HDL-C.
|
|
![]() |
C-reactive protein (CRP) |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Therapeutic Lifestyle Changes (TLC) |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
Diet is one of the essential elements for TLC. Despite recommendations to the contrary, the role of diet in the treatment of hypercholesterolemia has been largely neglected with the advent of the statins. The potential of diet to prevent CV disease (CVD) is often underappreciated by patients, who would rather take a pill than change ingrained habits, and by physicians, who perceive diet as ineffective and unimportant. Diet favorably alters the lipid/lipoprotein profile (Table 3) and may allow use of lower drug doses, with a reduced potential for adverse effects. In addition to reducing fasting lipid levels, diet may also reduce the risk of CHD by modifying atherogenic postprandial lipid responses that are not reflected in fasting values. Diet can help control or modify other CHD risk factors such as hypertension, obesity, and type 2 diabetes. Although difficult to implement, the role of diet is particularly important in children and young adults and in those in whom lipid-modifying drugs are intolerable or unacceptable.
|
|
Diet: major nutritional components
Saturated fatty acids (SFA). SFA decrease LDL receptor expression and increase LDL-C levels (10), whereas the long chain fatty acid, stearic acid, has little effect (17). For every 1% increase in total calories from SFA, the sum LDL-C increases by 2% (10). High-fat dairy products, meats, tropical oils, and hard shortening are major sources of saturated and trans fatty acids (TFA). Soy protein may be used to reduce animal food products, thereby decreasing SFA intake and leading to some small further reduction of LDL-C (18, 19). The ATP III recommends that SFA be less than 7% of total calories compared with the average United States adult intake of 11% (10).
TFA. TFA also decrease LDL receptor expression, elevate LDL-C, raise TG, and lower HDL-C concentrations (10, 20, 21). Conversion of liquid vegetable oils to solid or semisolid preparations, for the convenience of having stick or tub margarine, rehydrogenates the unsaturated fatty acids in the trans position. The greater the TFA content of these products, the more their effects resemble those of butter. Products containing hydrogenated vegetable oils are the major source of TFA. Baked goods such as crackers and cookies, and fried foods (especially when oil is used for frying), are high in TFA. Lesser amounts of TFA are found in animal sources, including dairy products. TFA intake represents almost 2.7% of total caloric intake in the United States (10), but should be decreased when possible because of its adverse effects on LDL-C, HDL-C, and TG (22). TFA may be associated with increased lipoprotein (a) [Lp(a); Ref. 20 ] and promote insulin resistance (21).
Dietary cholesterol.
Dietary cholesterol increases serum cholesterol to a lesser extent than SFA; 100 mg of dietary cholesterol increases serum cholesterol approximately 10 mg/dl per 1000 kcal (10). Daily average cholesterol intake is higher for men (331 mg) than women (213 mg). Foods rich in cholesterol include eggs, animal and dairy, poultry, and shellfish. Sterols in shellfish and shrimp do not appreciably influence the serum cholesterol (23, 24) unless fried, cooked in butter, or consumed in large quantities. Dietary cholesterol consumption should be less than 200 mg/d to maximize LDL-C lowering (Table 3).
MUFA and PUFA. When MUFA are isocalorically substituted for SFA, but the CHO content is not altered, the reduction in LDL-C is comparable to that observed with a high CHO-low fat diet, but there is little or no decrease in HDL-C or increase in triglyceride (13). Substitution of unsaturated for saturated fat favors CHD risk reduction (22) and may improve insulin sensitivity (20, 21). Although substitution of PUFA for SFA decreases CHD risk, there are no large populations that consume high amounts of PUFA (10). The relative value of dietary MUFA vs. PUFA in preventing CHD is unknown (25). However, there are differences. PUFA may lower LDL-C more (10, 20), has antithrombotic effects, and is antiarrhythmic (26), but theoretically may be more susceptible to oxidation (27). One of the primary sources of foods relatively high in MUFA, olive oil, also contains polyphenols and antioxidants (17).
Omega-3 fatty acids.
Omega-3 fatty acids derived from marine sources occur as docosahexanoic acid (DHA) and eicosopentanioic acid (EPA). Also, -linolenic acid, available from plant sources, can be elongated to EPA and DHA. Most prospective studies have reported that omega-3 fatty acids are associated with reduced sudden cardiac death or CHD mortality (26). Two-year mortality among CHD patients advised to eat fish two times per week or consume fish oil was decreased by 29% (28). One gram per day of fish oil reduced total deaths by 14% and CV deaths by 12% (29). Cardiac events were reduced by 5070% (depending on outcome measures used) in patients randomized to a Mediterranean diet and an
-linolenic acid supplement, independent of changes in blood lipids (30).
The protective benefits of omega-3 fatty acids are attributed to prevention of arrhythmias, decreased thrombogenicity, improved endothelial function, and reduced triglyceride levels (26). Only 1 g fish oil per day provides cardioprotection, but 35 g/d (23 g/d EPA and DHA) reduce TG by up to 30%, due to a reduction in VLDL production (26). Doses of this magnitude are associated with gastrointestinal adverse effects and may be difficult to sustain. The benefit of fish consumption is well known, and the American Heart Association recommends fish as part of a CHD risk reduction diettwo or more servings per week (31).
CHO. CHO substitution for SFA leads to a decrease in HDL-C and an increase in triglyceride (13, 32). This response is attenuated when complex CHO primarily consist of whole grains. CHO intake of at least 50% worsens lipid and nonlipid risk factors in patients with the metabolic syndrome. Current dietary recommendations emphasize complex CHO such as cereals and whole grains rather than simple sugars and starches. However, some CHO commonly referred to as complex have a higher glycemic index (GI) and consequently increase insulin responses more than simple sugars (33). Foods with a high GI have a larger increase in blood glucose for a given amount of CHO than do foods with a low GI. A high glycemic load (product of GI and CHO content) is associated with higher fasting triglyceride and lower HDL-C levels (33). Accurate classification of foods is difficult because the GI is affected by the cooking process, the rate of ingestion, and the consumption of other nutrients. Consequently, the GI does not easily lend itself to be a major factor in the diet component of TLC.
Plant sterols/stanols. Sitostanol/sterol esters inhibit intestinal absorption of dietary and biliary cholesterol and are most effective in persons with high cholesterol absorption and low cholesterol synthesis. Patients with apolipoprotein (apo) E4 have greater intestinal cholesterol absorption and may respond more favorably (20). Two to 3 g/d of plant-derived stanol/sterol esters reduces LDL-C by 615% without altering HDL-C or triglyceride (15). Plant stanols/sterols are available in grocery stores as regular and low-fat margarines (Benecol; Take Charge) at reasonable prices.
Vitamins/antioxidants.
Folic acid and vitamins B6 and B12 influence the metabolism and blood levels of homocysteine (21, 34). Homocysteine has atherothrombotic properties that include but are not limited to endothelial injury and/or promotion of thrombosis. Homocysteine has not been consistently related to CHD in observational studies (35). Yet, until data from clinical trials are available, it is reasonable to recommend folic acid and vitamin supplementation for those with elevated homocysteine levels who are at high risk for or with CHD (10, 34). Total plasma homocysteine concentration ranges from 515 µmol/liter in healthy adults (34). Generally, levels in men are higher than in women, and values increase with age in both sexes (36). Other factors that increase homocysteine levels include physical inactivity, smoking, excessive coffee or alcohol intake, and a variety of drugs including niacin and fibrates (36). Folic acid supplements attenuate the increase in homocysteine levels in patients taking fenofibrate (37). Homocysteine levels above 12 µmol/liter are thought to increase CHD risk. Approximately 20% of CHD patients have elevated homocysteine levels (5). Folic acid fortification of cereal grains has lowered homocysteine levels and the prevalence of elevated homocysteine levels (38). Although vitamins B6 and B12 may enhance the response, optimal doses are unknown, but recommended dietary allowances (RDA) are approximately 1.7 mg/d and 2.4 µg/d. If vitamin B12 deficiency has been excluded, a daily dose of 1 mg folic acid, 25 mg vitamin B6, and 0.5 mg vitamin B12 may be needed for high-risk patients with elevated homocysteine levels (34). Prospective randomized control trials of folic acid, B6, and B12 to reduce CVD risk are in progress (10).
Oxidative stress is an important factor for the development and progression of CHD (39). An inverse relationship exists between dietary antioxidants and CHD in epidemiological studies, whereas laboratory studies provide extensive support for the role of oxidative stress as an early and pivotal event in atherosclerosis. ß-Carotene and -tocopherol protect the LDL particle from oxidation, whereas ascorbic acid regenerates reduced
-tocopherol. Consequently, antioxidants are a logical choice to reduce the risk for atherosclerosis. However, the totality of evidence shows no relevant effects of antioxidants on the risk of CV events (5, 40). In the Cambridge Heart Antioxidant Study (CHAOS), 400800 mg of vitamin E daily in postinfarction patients reduced nonfatal myocardial infarction by 71%, but total mortality and CV deaths were increased in the treatment arm. The
-Tocopheral ß-Carotene Cancer Prevention Study and the (CARET) study found no CHD benefit from these antioxidants. Larger randomized trials have also failed to show any benefit from vitamin E. In the GISSI trial, antioxidant vitamin supplementation of patients with a recent myocardial infarction was of no benefit, whereas omega-3 fatty acid supplementation reduced cardiac events by 1015%. In addition, no differences in CHD outcome were seen in the Heart Outcomes Protection Evaluation Study (HOPE), which also used a factorial design in which half of the patients received an antioxidant vitamin supplement with or without ramipril (41). No CHD benefit was demonstrated in the large Heart Protection Study (HPS), in which half of the patients received antioxidant vitamin supplements, with or without simvastatin (42). Antioxidant vitamins partially inhibit the increase in HDL-C that occurs with niacin, inhibit regression of coronary atherosclerosis, and reduce clinical benefit in patients receiving a combination of niacin plus simvastatin (43). Consequently, there is no basis for recommending antioxidant vitamin doses in excess of the recommended dietary allowance. It may be that a lifetime of consuming a diet rich in antioxidants is different from relatively short-term consumption of antioxidant vitamin supplements and/or there are other unrecognized constituents of an antioxidant-rich diet that are cardioprotective or there are other favorable protective lifestyle elements in persons consuming a diet rich in antioxidants.
Alcohol.
There is a u-shaped or j-shaped relationship between alcohol consumption and CHD mortality (44, 45). Moderate alcohol consumption, defined as one drink per day for women and two for men, is associated with protection against CHD. One drink is equivalent to 5 ounces of wine, 12 ounces of beer, or 1.5 ounces of 80-proof whiskey. Potential protective mechanisms include an increase in HDL-C and apo A-I, a favorable influence on clotting factors, and reduced levels of inflammation (46, 47, 48). Inhibition of synthesis of endothelin-1, a vasoactive proatherosclerotic peptide, may also be beneficial (49). A recent large study reported no association with type of beverage, but rather with frequency of consumption. Three or more drinks per week were associated with approximately 30% reduction in risk of myocardial infarction (50). There continues to be controversy over whether the type of alcoholic beverage consumed makes a difference. Although most studies find no difference in the type of beverage and the reduction in CHD risk, there is considerable rationale to think that the constituents of red wine also have cardioprotective properties. Red wine contains phenolic compounds that inhibit lipoprotein oxidation and stimulate endothelial nitric oxide synthase expression (51); they may also be present in raisins and grape juice. The adverse effects of alcohol on blood pressure, cardiac rhythm, and myocardial function as well as its association with upper GI tract cancers must be considered. In addition, alcohol may aggravate preexisting hypertriglyceridemia and predispose to acute pancreatitis (48). The issue of alcohol consumption for CHD protection is obviously a very individual matter.
Other nutrients and supplements.
Numerous products have been recommended for lipid lowering and prevention of CHD. These include soy protein, isoflavones, nuts, and garlic (19, 44, 52). Preparations from health food stores are unregulated, and the bioactive content of these substances can vary within and among brands. In addition, there are increasing concerns about safety as well as efficacy. It is important to establish whether patients are using such products because of potential interactions with other drugs. No herbal or botanical substance has yet been proven to reduce the risk for CHD.
Soy proteins contain all of the essential amino acids and can be used as a substitute for animal protein, thereby reducing the SFA and cholesterol content of the diet. Twenty-five to 50 g/d soy protein is safe and can reduce LDL-C by 48% (18). This reduction may be greater in the presence of hypercholesterolemia. The U.S. Food and Drug Administration (FDA) has approved a health claim that daily consumption of 25 g soy protein as part of a diet low in saturated fat and cholesterol may reduce the risk of CHD (19).
Consumption of nuts is consistently associated with an inverse CHD risk in prospective studies (44, 53, 54). Most nuts are associated with changes in lipids/lipoproteins that reduce CHD risk (54). This association may be related to the MUFA, PUFA, and various constituents in nuts, such as vitamin E, arginine precursors, flavonoids, and other polyphenols (17, 55).
Aged garlic extract can reduce total and LDL-C by 510% in hypercholesterolemic patients (44). Most studies of garlic preparations have been poorly controlled, short in duration, and without standardization of the type of garlic preparation, making evaluation of efficacy difficult.
Weight loss
Most patients underestimate how much they eat and overestimate how much they exercise. Weight reduction must be an integral part of the dieting program for overweight and obese patients. However, ATP III recommends delaying weight reduction to avoid overloading patients with a variety of dietary recommendations (10). There are two phases to weight reduction: 1) early hypocaloric weight loss, in which the most marked lipid changes occur; and 2) long-term isocaloric weight maintenance, in which LDL-C and HDL-C levels return toward baseline. A reduction in body weight of 510% reduces total cholesterol by up to 18%, TG up to 44%, and LDL-C up to 22%, and increases HDL-C by up to 27% (56, 57). Weight loss also influences other major risk factors such as hypertension and type 2 diabetes. A recent randomized trial compared an Atkins-type hypocaloric diet containing less than 20 g/d of CHO (supplemented with fish, -linolenic acid, and flaxseed oil) to a hypocaloric high CHO-low-fat diet (58). There was greater weight loss, greater increase in HDL-C, and a greater reduction in TG over 6 months with the low-CHO ketogenic diet. Such a diet may be useful for the metabolic syndrome, but potential adverse effects, the need for supplements, and the difficulty in maintaining such a diet probably limit its effectiveness. A hypocaloric ATP III diet should be the cornerstone for weight loss. Only when such diets fail should alternative diets be considered.
Exercise
HDL-C is increased in a dose-dependent fashion with increased energy expenditure (59, 60, 61). The change in HDL-C is related to the distance run per week, time spent exercising, and possibly the frequency of exercise (61, 62). Twelve weeks of training at energy expenditure levels of 12002200 kcal/wk are usually needed to increase HDL-C. Physical activity reduces TG, more for those with higher baseline levels who have been inactive and minimally for those with relatively normal levels (10). Weight loss combined with favorable dietary changes enhances exercise-induced changes in the lipoprotein pattern, particularly in those with metabolic syndrome.
Physical inactivity is a major risk factor for CHD and substantially contributes to the lipid and nonlipid risk factors of the metabolic syndrome. Scheduled, regular exercise promotes energy expenditure that is necessary as our national economy continues to evolve from labor to service intensive and to maintain desirable body weight (body mass index, <25 kg/m2). The efficacy of resistance exercise for altering blood lipids is unresolved, but it may attenuate the postprandial TG response and decrease the fasting TG level (63). Postprandial lipemia can be reduced by a single aerobic exercise session performed 24 h before a high-fat meal; the reduction is related to the energy expended (62). Despite some inconsistencies, there is extensive evidence that aerobic exercise increases the activity of lipoprotein lipase, reduces TG, and increases HDL-C (62). A reduction in LDL-C mainly occurs in conjunction with weight loss (62, 63). Other possible mechanisms for the lipid changes are decreases in cholesterol ester transfer protein and hepatic lipase activity. Regular exercise improves insulin sensitivity in the absence of weight loss (62).
![]() |
Drug Therapy |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
Hydroxymethylglutaryl coenzyme A (HMG CoA) reductase inhibitors (statins). Statins are the drugs of first choice for patients with high or less than optimal LDL-C levels (10). They inhibit the rate-limiting step of cholesterol synthesis, create a transient decrease in intracellular cholesterol, increase the synthesis of the cell surface LDL receptor, and accelerate removal of LDL-C and triglyceride-rich lipoprotein. The latter accounts for the associated modest reduction in triglyceride levels that is observed with the use of these drugs, although more potent statins also inhibit VLDL synthesis. A mild increase in HDL-C, also occurs. Statins are the most potent, most effective, and best tolerated drugs for reducing LDL-C. In fact, availability of this class of drugs made it possible to achieve LDL-C reductions that reduced CV, and all cause mortality and provide confirmation for the lipid hypothesis.
Five statins are currently approved by the FDA for cholesterol lowering. Only three, (lovastatin, pravastatin, and simvastatin) are also approved for prevention of clinical CV events. Although it is likely that all drugs of this class will prevent coronary events, only these three have at this date been shown to do so in RCTs.
The cardioprotective property of these drugs is primarily due to reduction of LDL-C. It is also clear that these drugs have other effects, referred to as pleiotropic properties that may also be important (64). These pleiotropic properties importantly influence the biology of atherosclerosis by modulating immunoregulation, inflammation, coagulation, and vasomotor responsiveness and appear to do so independently of changes in LDL-C. Whether all drugs of this class have similar or identical pleiotropic properties has not been systematically investigated. A recent study of 40 mg/d pravastatin vs. placebo in high CHD risk patients raises serious questions about the cardioprotective pleiotropic properties because LDL-C reduction was modest and there was no decrease in CHD events (65, 66). Several recent studies have demonstrated cardioprotective benefits of statins in persons with low or near optimum LDL-C, suggesting that statins may be antiatherogenic agents (42). The decision to use a statin is still based on the need for LDL-C reduction and not on putative pleiotropic properties.
Despite their efficacy in large primary and secondary prevention trials, statins have been associated with a maximum CHD risk reduction of 35% (7). It is quite likely that the residual risk for CHD events is partly related to suboptimal lipid modification; untreated or inadequately treated traditional CHD risk factors such as diabetes, hypertension, cigarette smoking, and dyslipidemia; and/or to emerging risk factors such as Lp(a), small LDL particles, homocysteine, and insulin resistance whose role has not been clearly defined.
The statins vary in potency, physicochemical characteristics, and routes of metabolism. They are among the safest drugs developed, and most patients tolerate them without serious adverse effects. A few patients may develop serious complications. In many large clinical studies, the risk of hepatic toxicity or myopathy is no greater than that observed in patients receiving placebo. The American Heart Association, American College of Cardiology, and the NCEP have recently reviewed these issues and concluded that statins are safe for the vast majority of patients who take them (67). Why a few patients develop serious complications is unclear. However, concomitant use of drugs that are metabolized by the same drug pathway is clearly one important cause for toxicity. In addition, the presence of other serious complex medical conditions, the use of excessive doses of statins in patients on numerous other medications, or the aggressive use of statins in patients with renal or hepatic disease may explain some of the toxicity (68). Also, there are probably pharmacogenetic reasons for susceptibility to drug toxicity. It is also clear, that patients included in clinical trials are fundamentally healthier, and the low toxicity in trials may not accurately reflect the risks of many of the patients that are seen in everyday practice. The issue of myopathy is complicated by the fact that patients treated with statins can have increases in creatine kinase (CK) without myalgia, have myalgia with no increase in CK, or have myalgia with an increase in CK. Recently, a small group of patients with muscle symptoms and normal CK were demonstrated to have histological evidence of myopathy (69). Many of us have encountered patients with myopathic symptoms and normal CK whose symptoms disappeared when the drug was stopped and reappeared when it was restarted. In addition, some patients have symptoms when they are switched to other statins, whereas others experience symptom relief. The significance of all of this is unclear.
Selection of a statin should be based on the extent of LDL-C reduction that is desired, cost, and whether the selected statin has been shown to reduce clinical CHD events. The concept that lower is better (70), although likely, has not been proven, yet it has been endorsed by most physicians because atorvastatin accounts for at least 60% of the statin market (71). Although it is probable that atorvastatin will be cardioprotective, no large RCT has yet been published demonstrating its efficacy in reducing CHD events. It is also curious that most physicians choose atorvastatin because of its potency but never titrate the dose to take advantage of this property, and many also erroneously believe that atorvastatin is a good first drug for patients with hypertriglyceridemia. All statins are capable of reducing triglyceride, but triglyceride reduction parallels LDL-C reduction (72). The greater the reduction in LDL-C, the greater the reduction in triglyceride. Patients with hypertriglyceridemia often have modest hypercholesterolemia, which can be due to cholesterol in VLDL or LDL. LDL-C cannot be calculated when the triglyceride is more than 400 mg/dl. For patients with increased LDL-C and triglycerides of 200400 mg/dl, atorvastatin or simvastatin are the drugs of choice. Maximum triglyceride reduction in this setting is approximately 30% with atorvastatin and approximately 20% with simvastatin. Modest increases in HDL-C occur with statins, but the effect is attenuated or abolished when large doses of atorvastatin are used (73, 74).
Statins, except atorvastatin, are usually dosed at night because of higher nocturnal cholesterol synthesis. Studies suggest slightly greater efficacy. Statin tablets can be cut when attempting to curtail expense, but only one tablet at a time should be cut so that the unused half can be taken the next day. This is particularly important in the elderly where drug costs are a vexing problem, but this may be less of an issue for products where there is level pricing. Statins have also been used on alternate days with a small reduction in efficacy. Alternate day dosing is not desirable because it is difficult for many patients to remember to take their medication.
The efficacy of a selected dose of statin can be observed within 34 wk so that titration of the dose can be accomplished rapidly. It is also true that the effect is rapidly lost when the statin is stopped. The reduction in LDL-C is 3035% when equivalent doses of statins are used (10 mg atorvastatin = 20 mg simvastatin = 40 mg lovastatin/pravastatin = 80 mg fluvastatin XL; Ref. 70). Each doubling of the dose further reduces LDL-C by 6%. Doubling the dose has no financial impact where there is level pricing. The benefit of dose doubling is relatively small, and it should be remembered that toxicity is directly related to dose. When the degree of LDL-C lowering can be achieved by increasing the dose, the question should be asked whether this is the safest and most cost-effective way of further reducing LDL-C. For example, use of a stanol/sterol margarine can reduce LDL-C by 1012% at a fraction of the cost and at much lower potential toxicity than doubling or quadrupling the statin dose. More recently, the addition of ezetimibe, a new cholesterol absorption inhibitor, has been shown to reduce LDL-C by approximately 20% when added to statin therapy (75). The expected LDL-C reduction with any given statin dose represents an average of patients taking that dose. Because some patients are sensitive and others are resistant, it is impossible to know how an individual patient will respond. For reasons that are not well understood, some patients will be resistant and experience a reduction of less than 10%. The efficacy or lack of efficacy may in part be related to the efficiency of intestinal cholesterol absorption. Those whose hypercholesterolemia is primarily exogenous due to efficient cholesterol absorption will be less responsive, whereas those with endogenous hypercholesterolemia will be sensitive (76). This lack of responsiveness appears to extend to all statins and cannot be circumvented by use of a different statin. A gradual loss of efficacy (tachyphylaxis) has also been described for some patients (77). A short drug holiday and then reinitiation of therapy has been associated with restoration of the expected response.
Bile acid sequestrants. These agents have been available for years and in that regard are time honored. They bind bile acids in the intestine, preventing their reabsorption in the terminal ileum, and reduce the hepatic pool of bile acids that leads to increased intracellular conversion of cholesterol to bile acids in hepatocytes. This transient decrease in intracellular cholesterol leads to a compensatory increase in HMG CoA reductase activity, increased synthesis and expression of the LDL receptor, and a subsequent increase in the catabolism of LDL particles. The efficacy of this class of drugs is offset by the increase that occurs in HMG CoA reductase and in cholesterol synthesis. However, use of a statin in combination with a sequestrant prevents the increase in cholesterol synthesis, thereby further augmenting the decrease in LDL-C. The cholesterol reduction that occurs with this combination is the sum of their individual effects (78). Sequestrants increase VLDL synthesis and the serum triglyceride levels to a variable degree and may exacerbate or cause hypertriglyceridemia, particularly those with borderline high TG. A paradoxical increase in total cholesterol that may be observed in patients treated with sequestrants is due to increased synthesis of VLDL and VLDL cholesterol (79).
Bile salt binding resins are inconvenient to use because they have to be mixed with liquids or foods; however, colestipol is also available as a pill. Palatability and adverse gastrointestinal effects of sequestrants limit use of large doses. The reduction in LDL-C with maximum doses is approximately 30%, but a significant reduction in LDL-C, of the order of 15%, can be achieved by using two scoops per day (78, 80). Sequestrants are clearly safe and effective cholesterol-lowering agents and do not interfere with the absorption of fat soluble vitamins when low to modest doses are used but may interfere with the absorption of various medications. Their primary role at this time is as an adjunct to statin therapy or as monotherapy for children with hypercholesterolemia. Eventually, they will probably be replaced by ezetimibe.
A relatively new preparation, colesevelam, a soluble fiber product, is also available. Its advantage is that it is more convenient to use because it is available as a pill rather than having to be mixed in juice or other liquids. This preparation can reduce LDL-C by 1015%, and its effects are additive to those of statins (70). Side effects are similar to those of resins, but may be less frequent. It is expensive and is unlikely to be the drug of first choice unless use of other drugs is contraindicated or unacceptable.
Cholesterol absorption inhibitors [ezetimibe (Zetia)]
The first drug of this class was approved by the FDA in November 2002. It interferes with the absorption of cholesterol, and when used as monotherapy, reduces LDL-C by 1520% (75). It is effective in patients on a low SFA, low cholesterol diet because it blocks reabsorption of cholesterol secreted into bile and the enterohepatic circulation of endogenously produced cholesterol. Endogenous cholesterol synthesis is approximately 900 mg/d, several times greater than cholesterol intake. The reduction in LDL-C is due to increased endogenous catabolism of LDL. Its effects on triglyceride and HDL-C levels are trivial. It can be used very effectively in combination with a statin (81, 82). Ezetimibe provides approximately 70% of the reduction expected in LDL-C when compared with the first recommended dose of a statin and three times the response expected by doubling the statin dose. It should be considered as an adjunct to therapy in patients where LDL-C reduction is suboptimum. Ezetimibe is recommended once daily at a dose of 10 mg. Its effects on CV endpoints are entirely unknown.
Fibrates. Although four fibrates are available worldwide, only three fibrates are available in the United States, and of these, only two (gemfibrozil and fenofibrate) are of any consequence. Fibrates are most effective for treating patients with hypertriglyceridemia and reduced HDL-C (10, 70). The efficacy of these drugs for reducing triglyceride and increasing HDL-C is related to the magnitude of the hypertriglyceridemia. However, the higher the pretreatment triglyceride level, the less likely that it will be normalized by fibrate treatment. Whether fenofibrate and gemfibrozil substantially differ in their lipid/lipid protein effects is not known because there have been few properly designed head-to-head trials. Gemfibrozil has a modest LDL-C-lowering effect in patients with hypercholesterolemia, has little effect on LDL-C in patients with mixed hyperlipidemia, and increases LDL-C in patients with pure hypertriglyceridemia (83). Its ability to increase HDL-C in patients with hypertriglyceridemia is impressive, but minor changes occur in patients with low HDL-C and normal triglycerides (84). Fenofibrate may be slightly more effective than gemfibrozil in reducing LDL-C in patients with hypercholesterolemia or mixed hyperlipidemia (85, 86). Despite mild to modest effects on LDL-C, both drugs cause a desirable shift in LDL particle size from small/dense (pattern B) to large/buoyant (pattern A; Refs.87, 88, 89, 90). Gemfibrozil reduced CHD events by 35% in the Helsinki Heart Study (83) and by 22% in the Veterans Administration Low High Density Intervention Trial (91). Fenofibrate reduced progression of coronary atherosclerosis in the Diabetes Atherosclerosis Intervention Trial, as well as reducing clinical CV events, but the study was underpowered to evaluate its effects on clinical endpoints (92). There are no data on the effectiveness of these drugs in women.
Drugs of this class reduce triglyceride and increase HDL-C by stimulating peroxisome proliferator activator receptor (93). The reduction in triglyceride is mediated by enhanced clearance of triglyceride-rich lipoprotein and decreased VLDL synthesis. Changes in HDL-C and the size of the LDL particle may be secondary to reduced triglyceride levels and/or to increased synthesis of apo A-I and apo A-II. In the Veterans Affairs HDL Intervention Trial (VA-HIT), the increase in HDL-C of 2 mg/dl accounted for only approximately 20% of the reduction in CV events produced by gemfibrozil (94). Nuclear magnetic resonance lipoprotein analysis of samples from the VA-HIT indicate that a decrease in LDL particle number and an increase in HDL particle number accounted for approximately 80% of the reduction in CV events (95). There was no change in LDL-C and a very small change in HDL-C. Therefore, qualitative lipoprotein changes produced by gemfibrozil rather than quantitative changes may be important. However, several caveats must be mentioned: nuclear magnetic resonance lipoprotein analysis is still an experimental technique, and post hoc analyses can only adjust for confounding variables that are known and cannot adjust for those that exist but are unknown. Whether fibrates also have additional pleiotropic properties (93), like statins or thiazolidinediones (TZDs), is an important question.
Fibrates are usually well tolerated but can be associated with adverse effects (hepatic, gastrointestinal, rash) and an increased risk of gallstones. An increased risk of gastrointestinal malignancy was observed in the World Health Organization Clofibrate trial, but not in the Helsinki Heart Study or VA-HIT. Myopathy can occur with fibrates, usually in patients with end-stage renal disease (ESRD). There are numerous reports of severe myopathy and rhabdomyolysis in patients treated with fibrates (mainly gemfibrozil) and statins, but the combination is relatively safe with an incidence of myopathy estimated at 12 per 10,000 (96).
Fixed doses of fenofibrate and gemfibrozil are used; there is no evidence that dose titration is valuable or effective. Because fibrates are cleared by the kidneys, the doses should be reduced by 50% when fibrates are used in ESRD. When fibrates are added to statin therapy, it is advisable to reduce the statin dose to 2550% of the maximum recommended dose to minimize the risk of myopathy. Because there is some additional risk of using fibrates in combination with statin, their use should be restricted to patients with or at high risk of CHD where the benefit exceeds the risk of adverse effects.
Niacin (nicotinic acid). Niacin is a drug that reduces Lp(a), LDL-C, and triglyceride and increases HDL-C and LDL particle size, all desirable antiatherosclerotic changes (97, 98, 99). Large doses must be used, which has adverse effects preventing the use of niacin for some patients or requiring the use of lower total daily doses for others. It is available in immediate-release and delayed-release preparations (98, 99). Its major adverse effects are hot flashes, pruritus, gastric irritation, hepatotoxicity and precipitation or worsening of glucose metabolism. Niacin also increases serum homocysteine and uric acid levels. Nicotinamide, although well tolerated, cannot be substituted because it has no lipid-lowering properties. Some patients tolerate the immediate-release preparations (approximately 50%), and a greater percent (85%) tolerate the delayed-release preparations (98). However, even among tolerant patients, there are times when hot flashes and burning occur for unexplained reasons. A variety of tricks are used to reduce the unpleasant skin sensations: use of an aspirin 30 min before taking the niacin as well as avoiding hot baths/showers, hot liquids, and alcohol are helpful. Immediate-release niacin is cheap and wherever possible should be used. Very large doses of immediate-release niacin have been used in the past, but the practice now is to use smaller doses, particularly when it is added to other lipid-lowering drugs. The maximum dose of extended-release niacin should not exceed 2000 mg/d, to avoid hepatotoxicity, and doses of 10001500 mg/d are most reasonable, particularly because its role is to supplement or amplify therapy with a LDL-C lowering agent (10). As a general rule, lower doses are required for increasing HDL-C than for reducing LDL-C and triglyceride levels (99). This may be important when the problem being addressed is isolated low HDL-C. Niacin has been used very effectively in combination with statins (100) as well as with bile acids sequestrants (70). Use of 40 mg lovastatin in combination with 2000 mg of extended-release niacin is well tolerated; it reduces LDL-C by approximately 30% and triglyceride by approximately 35% and increases HDL-C by approximately 35% (100). Although nicotinic acid increases insulin resistance and can precipitate hyperglycemia, it has a relatively small adverse impact on glycemic control in patients with established type 2 diabetes mellitus and modest hyperglycemia that can be prevented or minimized by adjusting their hypoglycemic medication (101, 102). Myopathy has been reported when used in combination with statins, although this occurs even less frequently than with fibrates. When statins alone have not reduced LDL-C to goal and there is coexistent dyslipidemia or when isolated low HDL-C is being treated, niacin is an excellent choice. Fibrates are probably preferable as the first drug for combined hyperlipidemia when the triglyceride is at least 400 mg/d. Niacin may be used in combination with fibrates for hypertriglyceridemia that is resistant to monotherapy.
TZDs. Questions are commonly asked whether this class of drugs has important effects on lipids/lipoproteins and whether one drug is better than another. These are important issues because it is likely that physicians are or will be using these drugs to prevent or treat diabetes mellitus in patients with dyslipidemia. The stimulus to therapeutic intervention with these drugs in patients without diabetes is due to the large numbers of patients with the metabolic syndrome and its association with increased morbidity and mortality in men and women (103, 104). Virtually all data on lipids and TZDs come from retrospective observational studies. Pioglitazone appears to be more effective in reducing triglyceride and LDL-C and increasing HDL-C than rosiglitazone, although the effects on glycemic control are comparable (105). However, quantitative differences in LDL-C and HDL-C changes produced by the drugs are small at best. Comparison of lipid profiles in patients switched from troglitazone to rosiglitazone suggest that rosiglitazone increases LDL size and HDL2 compared with troglitazone (106). There are no head-to-head trials of pioglitazone and rosiglitazone on quantitative and qualitative changes in lipids and lipoproteins. Furthermore, despite putative pleiotropic properties that influence the basic molecular events of the atherosclerotic process (107, 108), there are no data that demonstrate that drugs of this class reduce CHD events.
Treatment paradigms
The recommendations below are in addition to TLC or when immediate therapy is necessary.
Hypercholesterolemia (increased LDL-C). Statins are the drugs of choice for this disorder. Although each individual will have a unique response, drug selection should be based on the magnitude of the reduction desired to reach specific goals. Doses should be titrated at 3- to 4-wk intervals, recognizing that doubling the dose further reduces LDL-C by only 6%. At some point, it becomes more cost-effective and perhaps safer to add a second drug then to increase the statin dose. Statins should be started immediately in all patients with symptomatic CHD (ACS, angina) and simultaneously with lifestyle modification in patients with established CHD (10, 109, 110). In fact, studies suggest benefit in patients with or at high risk of CHD even when LDL-C is less than 100 mg/dl (42). This suggests that statins have antiatherosclerotic properties and should be used in patients with atherosclerosis regardless of LDL-C levels. Modest further reductions in LDL-C on the order of 10% can be achieved by adding plant stanol/sterol margarine, 15% by adding bile acid sequestrants, and approximately 20% by adding ezetimibe. Lower is probably better, but efficacy and safety have not yet been prospectively proven in RCT designed to address this question. There is no proven adverse effect(s) of LDL-C below 70 mg/dl, and newborns are healthy with LDL-C levels of 3050 mg/dl. Because further lowering of LDL-C may be of diminishing returns, treatment of coexistent lipid disturbances, such as increasing the HDL-C concentration (111) or aggressively modifying other risk factors, may be more effective (112).
Mixed (combined hyperlipidemia). Achievement of optimum LDL-C targets takes precedent in this group of patients with multiple lipid/lipoprotein disturbances. Consequently, therapy should be initiated with a statin. Potent statins, such as atorvastatin, simvastatin, and rosuvastatin (when approved by the FDA) make the most sense because substantial triglyceride reduction can also be expected, but probably not exceeding 35% with current products. When pretreatment triglyceride is below 400 mg/dl, it is possible to calculate the LDL-C concentration. When the triglyceride is above 400 mg/dl, the LDL-C cannot be calculated. Although LDL-C can be directly measured in such patients, this is seldom done. Therapy can be initiated with a fibrate or niacin, although their effects on the lipoprotein profile differ. Gemfibrozil is better than niacin for reducing the triglyceride level but less effective than niacin for increasing HDL-C; niacin also reduces LDL-C, whereas gemfibrozil does not (84). When the triglyceride decreases to less than 400 mg/dl, calculation of the LDL-C permits a decision regarding subsequent statin therapy. After achievement of the LDL-C target, the non-HDL-C should determine further additional therapeutic decisions. Statin-fibrate and statin-niacin combinations are more effective than statins alone in correcting lipid abnormalities in combined hyperlipidemia (113, 114). The use of statins in combination with fibrates or niacin increases the risk of adverse events; consequently, the benefits of such therapy should clearly outweigh the risks. Because you cannot prevent what is not happening, the use of these combinations in persons with low short-term risk is not warranted. Adverse events from drug combinations occur equally across all risk strata, whereas benefits occur only in those most likely to have events.
Hypertriglyceridemia. Patients in this group are heterogeneous with regard to underlying mechanisms and the accumulation of triglyceride-rich lipoproteins that account for the hypertriglyceridemia. A variety of drugs and estrogen-containing preparations can cause or accentuate other genetic and acquired factors that may exist in these patients. Assuming that secondary causes can be identified and eliminated and that patients with increased chylomicrons can be identified, patients in this category have either familial hypertriglyceridemia or combined hyperlipidemia, where the increase in LDL-C has not yet been expressed. Patients with combined hyperlipidemia classically have multiple lipid phenotypes within their family and within themselves from time to time depending on other metabolic and dietary factors (115). Both groups have reduced levels of HDL-C and small LDL particles, but may be differentiated by increased apolipoprotein B and decreased apolipoprotein A-I levels in combined hyperlipidemia. Because triglyceride levels are often above 400 mg/dl, it may not be possible to immediately differentiate these disorders unless family data showing multiple lipid phenotypes are available. The concept that patients with familial hypertriglyceridemia are not at increased CHD risk no longer appears to be reliable (116).
Therapy in persons with hypertriglyceridemia should include extensive lifestyle changes, elimination of secondary factors, and the initiation of therapy with a fibrate. Insulin sensitivity is improved by weight loss and physical activity. In our experience, this seldom reduces triglyceride to optimum levels unless the initial level was only modestly increased and adjunctive measures may be required or desired. Obviously, the imperative to be more aggressive should be based on projected CHD risk, a persisting undesirable non-HDL-C level or other evidence, such as LDL size, that justifies intensified therapy. Omega-3 fatty acids can be used to reduce triglyceride levels, but their efficacy is dose related, and large doses create problems related to dosing and adverse effects that most patients will not tolerate, at least not for long, and convenience.
Isolated low HDL-C. Many patients with or at risk of CHD have low HDL-C as an identifiable CHD risk factor (111). Some patients with low HDL-C have hypertriglyceridemia and therefore are not isolated, whereas others do not. The relationship between CHD risk and HDL-C is the mirror image of that which exists with LDL-C. Consequently, increasing the HDL-C a small amount may be far more effective in reducing CHD than decreasing the LDL-C by a comparable amount when LDL-C is near or at goal (111). The only drug currently available that substantially increases HDL-C in patients without hypertriglyceridemia is niacin (98). The antiatherosclerotic properties of HDL-C are commonly attributed to reverse cholesterol transport, but it is very clear that HDL has other important cardioprotective properties (117). There must be a better understanding of these other properties of HDL to determine when or whether low HDL-C is a CHD risk factor and whether all drugs that increase HDL-C will be cardioprotective. Niacin is known to be cardioprotective (118) and to increase HDL-C and apolipoprotein A-I by selectively enhancing hepatic removal of the cholesterol ester from HDL by hepatic sterol receptor-B1 while not removing the HDL particle from the circulation (97). The ATP III does not provide specific goals for low HDL-C, but every milligram per deciliter helps. It is not unreasonable to attempt to increase HDL-C to more than 45 mg/dl (111), although this may be beyond reach for many patients with current medications.
Therapeutic implications of LDL size. There is great interest in the relationship of dense LDL (dLDL) to the risk of CHD and perhaps the response to therapy. Epidemiologic data and bench research support its importance as an emerging CHD risk factor (119, 120, 121, 122). Changes in dLDL with therapy are associated with CHD risk reduction, but no prospective randomized trials have been published that evaluate the effects of diet and/or pharmacological intervention on the relationship of changes in LDL to CHD events. The problem is confounded by the metabolic interrelationships between dLDL and plasma triglyceride and HDL-C levels, as well as associations with insulin resistance, the metabolic syndrome, cigarette smoking, sedentary lifestyle, and hypertension, to name a few. Multivariate statistical analyses cannot be used to determine which parameters are independent or dependent when they are metabolically intertwined. That approach trivialized the importance of serum triglyceride as a CHD risk factor for years because of its inverse relationship to HDL-C.
Whether knowing LDL size or obtaining sequential measurements, as recommended by several laboratories, improves therapy or the effectiveness of therapy is unclear. We believe that following the ATP III guidelines provides effective therapy for the vast majority of patients with or at risk of CHD, even those with dLDL. Statins are clearly superior to all other medications for reducing LDL-C concentration and particle number. They reduce the dominant LDL particle but generally do not change the LDL particle (122, 123, 124). However, this is not as clear as it seems because recently both fluvastatin and atorvastatin have been shown to shift the LDL profile toward more buoyant particles (125, 126, 127) but pravastatin and simvastatin have not (123, 124, 127). Treatment to goal is the primary objective in any patient. If goal is achieved, then the non-HDL-C should identify patients for additional therapy with fibrates or niacin (10). Both drugs reduce triglyceride and non-HDL-C and increase HDL-C. Neither drug is as efficient as statins in reducing LDL particle number, but both cause or are associated with an increase in LDL size that is presumed to be beneficial. One could argue that the target triglyceride should be reduced from 150 mg/dl to no more than 100 mg/dl to minimize the number of patients with dLDL (estimated to still be 50% with TG = 150 mg/dl; Ref. 128). For those who argue that manipulation of LDL size is important, it is equally unclear whether a statistically significant change in size is clinically significant. We think that a rare patient may benefit from measurement of LDL size, i.e. patients with a strong family history of CHD but no obvious risk factors or patients with CHD who continue to have events despite optimization of risk factors.
Special Patient Populations
Women CHD is the leading cause of death in women and is largely due to the risk factors that occur in men (129). The mortality rate for coronary events is equal to or greater than the rate for men (129). Development of CHD generally lags about 10 yr behind men, and perhaps longer for the more serious formssudden death and myocardial infarction. The ATP III recognizes age above 55 yr as a CHD risk factor for women (10). The increase in CHD with age occurs in a log-linear fashion even after menopause (130). The age-related increase in CHD after menopause has led to the concept that endogenous estrogen is protective. Several trials of postmenopausal women randomized to hormone replacement therapy or placebo have been published that do not support this concept (131). The Heart Estrogen Progestin Replacement Study demonstrated that estrogen-progestin therapy was no more effective in secondary prevention of CHD than placebo, despite favorable lipid changes (132). The Estrogen Replacement and Atherosclerosis (ERA) trial, which used estrogen-progestin therapy, estrogen monotherapy, or placebo, demonstrated no reduction in coronary atherosclerosis or clinical CV events despite beneficial lipid changes (133). The estrogen-progestin arm of Womens Health Initiative, a primary prevention trial involving 16,000 women, revealed a small but significant increase in coronary events and strokes after an average 5.1 yr of follow-up (9).
Data from randomized clinical trials consistently document the value of treating dyslipidemia in women (7, 134). The Program on the Surgical Control of the Hyperlipidemia, in which LDL-C lowering was achieved by partial ileal bypass, demonstrated that aggressive treatment of hyperlipidemia in women reduced CHD events (135). In the AFCAPS/TexCAPS primary prevention trial, CHD events were reduced slightly more in women (46%) than in men (37%; Ref. 136). In the Cholesterol and Recurrent Events (CARE) study, absolute risk reduction was greater for women, and the benefit tended to occur earlier than for men (137). In the Scandinavian Simvastatin Survival study, the reduction in coronary events among women was comparable to that for men (138). The HPS included more than 5000 women aged 4080 yr with a substantial 5-yr risk of death from CHD (42). Reduction of major vascular events in women who received 40 mg simvastatin was approximately 25%. In the Pravastatin in Elderly Individuals at Risk of Vascular Disease (PROSPER) study, women ages 7080 yr were as likely to benefit as men (139). In contrast, women in the LIPID trial had a nonsignificant 11% reduction in CHD events compared with a 26% reduction in men (140).
Lipid abnormalities in women should be treated as aggressively as in men. With data on postmenopausal hormone replacement therapy showing no protection against CHD, treatment of lipids appears increasingly important. The ATP III lipid goals are the same for women as for men (10). Special attention should be given to the important independent effects of triglyceride and HDL-C that appear to have a greater impact on CHD risk than in men.
Older patients Although age is the most powerful risk factor for CHD, there has been reluctance to use lipid-lowering medication for prevention of CHD in the elderly. Eighty-five percent of the CHD events occur in persons over 65 yr of age (129). Nearly 40% of those 65 yr and older, an expanding segment of the United States population, are eligible for treatment (129). Treatment has been questioned because of the belief that older persons are likely to have short life spans and advanced irreversible atherosclerosis that is unresponsive to treatment. However, the lag period for efficacy in the major statin trials is no more than 2 yr (141), and more recent studies of ischemia, endothelial function, and ACS (7, 142) indicate that benefits may occur within weeks to months of therapy, well within the life expectancy of healthy older persons (143). There are several reasons for the greater absolute CHD risk that occurs with age: major coronary risk factors increase dramatically with age, particularly systolic hypertension and the metabolic syndrome, and impaired glucose intolerance and type 2 diabetes mellitus increase in frequency. In general, aging is associated with an increasing atherosclerotic burden, and subclinical CHD is common. Approximately 60% of older persons and 88% of those with diabetes have CHD (144). Treatment of the elderly should be considered as secondary prevention because age is a major, albeit imperfect, surrogate for atherosclerosis.
The major clinical trials have conclusively demonstrated that older patients are as likely to benefit as younger adults from treatment of dyslipidemia (145, 146). The atherosclerotic process in older persons is not different from that of younger persons, nor has the response to lipid modification differed. Stroke is also an increasingly important problem among aging persons. All of the major statin trials and the VA-HIT have shown a reduction in atherothrombotic stroke with lipid lowering (92, 147, 148). This is particularly true for the secondary prevention trials. This is an important additional benefit for the elderly. In a study of ACSs lasting only 16 wk, 80 mg of atorvastatin decreased fatal and nonfatal stroke events (149). The risk reduction in stroke, especially fatal stroke, was 31% and was evident after only 612 months. Although of great importance, the stroke benefit reported with statins has been documented primarily in patients with CHD. Further data are needed to establish the role of statins in patients without documented CHD but with high risk of cerebrovascular disease. Because of the nature of primary prevention studies, there are only a small number of cerebrovascular accidents for individual trials. However, data from metaanalyses estimate that stroke risk (fatal and nonfatal) is reduced by 1520% in primary CHD prevention trials (147). The AFCAPS/TexCAPS trial demonstrated significant reduction in the rate of stroke from 1.36% in control group to 0.82% in the lovastatin group (odds reduction, 40%; Ref. 148). The HPS also had a large group of diabetics with a plasma total cholesterol greater than 135 mg/dl. During the 5-yr treatment period, allocation to simvastatin was associated with a significant 25% reduction in stroke. This was mainly because of a reduction in ischemic stroke (placebo, 4%; simvastatin, 2.8%; P < 0.0001) without any adverse effects of simvastatin on hemorrhagic strokes (42). However, the PROSPER (139) and ALLHAT (65) studies showed no effect of pravastatin on stroke risk among the elderly.
Treatment of lipid disturbances in the elderly is more challenging than in younger patients. TLCs, the first line of therapy, may be difficult to implement because of fixed habits and the restrictions imposed by age and health status on diet and exercise. Changes in nutrition and body composition may also affect pharmacokinetic properties of drugs and influence the response to lipid-lowering agents. Older Medicare patients often have financial problems that create dilemmas and tough decisions that might complicate the regular use of lipid-altering medication. The frequent use of multiple drugs, some of which alter lipids or interact with lipid agents, adds to the complexity of management.
Changes in renal and hepatic function with age and comorbidity influence adherence to therapeutic regimens as well as contribute to potential and real adverse drug interactions. Because of the coexistence of multiple medical problems, secondary causes of hyperlipidemia, including medications, must be excluded in older patients. Several common medical problems that may compromise therapy and predispose to adverse effects need to be addressed. The first is hypothyroidism, which leads to increased total and LDL-C due to decreased catabolism and also to decreased metabolism of statins that may predispose to myopathy. Elevated levels of CK may be present in patients with moderate to severe hypothyroidism before statin therapy. The second important medical condition that may cause adverse effects is renal insufficiency. The serum creatinine level represents the balance between creatinine production and glomerular filtration and creatinine excretion. In older patients, lean body mass (mainly skeletal muscle) is reduced, which leads to lower rates of creatinine synthesis and consequently serum creatinine concentrations that are inappropriately low relative to glomerular filtration rate. Glomerular filtration rate also progressively decreases with age so that a normal range serum creatinine level can be observed in elderly patients with moderate to marked renal insufficiency (7). This is probably of no consequence with statin therapy, but is a very important consideration when fibrates are used. Consequently, renal function should be assessed before use of fibrate monotherapy or fibrates in combination with statins.
Because age is an important surrogate for CHD, it is difficult to know whether a healthy elderly person with undesirable lipids is protected against CHD or on the threshold of an event. The presence of subclinical disease or perhaps a CRP level (12) as an indicator of risk may help resolve the dilemma. Older patients respond as well as younger patients to lipid therapy (145), and CVD benefits may extend life, improve quality of life, and reduce the use of health care resources. Postponing morbidity without prolonging life, the compression of morbidity hypothesis (150), is a very acceptable goal. Although fewer elderly patients have to be treated to prevent an event, suggesting greater cost efficacy, older patients have a shorter life expectancy, and quality life years gained will be shorter. There is good evidence that controlling lipid levels in persons at least 65 yr of age who are at increased CHD risk but in relatively good health is beneficial. Hence, there is no rationale for excluding older patients from treatment of lipids solely on the basis of age.
Children and young adults Early atherosclerotic lesions, or fatty streaks, begin in childhood and are related to cholesterol levels and other CHD risk factors that exist in adults (151). Progression of fatty streaks to fibrous arterial plaques is dependent on traditional CHD risk factors (152). Genetic disorders and environmental factors accelerate the childhood development of atherosclerosis. Aggressive prevention strategies are needed in youth because of the increasing prevalence of childhood obesity and type 2 diabetes. Undesirable levels of LDL-C and other lipids, present in childhood, often persist (track) into adulthood (153). Use of age-specific normal values of cholesterol, triglyceride, LDL, and HDL-C is required to identify and treat lipid disorders in children (153). Cholesterol levels should be monitored if a child has: positive family history of CVD (including premature CHD no later than age 55 yr in parents, grandparents, aunts, or uncles); parents with history of blood cholesterol levels above 240 mg/dl; other CVD risk factors; or no family history available. Physicians may elect to screen a child for high blood cholesterol if any of the following risk factors are present: hypertension, smoking, sedentary lifestyle, obesity, excessive alcohol intake, certain medications associated with dyslipidemia, or disease states such as diabetes mellitus or nephrotic syndromes (154).
TLCs should be initiated as soon as feasible. After 2 yr of age, the diet is gradually changed to an adult type of diet. This provides the opportunity to steer children toward a lifelong habit of more healthful food choices (155). Dietary effects on serum lipids in children are similar to those observed in adults (156). These diets are safe, and growth and development are normal (157). A coronary risk factor intervention project conducted in healthy children with normal lipids demonstrated a 56% reduction in total cholesterol with a diet restricted in total saturated fat without any adverse effects (158). Physical activity is critical for long-term weight maintenance and may be especially helpful in those children with hypertriglyceridemia and associated low HDL-C. Plant sterols and stanols with soy protein, as an alternative to meat products, to lower LDL-C should be considered as well (159). Short-term data indicate that children may safely ingest plant stanols/sterols. Nevertheless, with the potential for lowering of ß-carotene and other fat soluble vitamins, the effects of the supplements on children over prolonged periods of time need to be monitored.
Drug therapy should be started if cholesterol goals are not met by TLC and the LDL-C is at least 190 mg/dl or LDL-C is at least 160 mg/dl, with a positive family CHD history, premature CHD, or at least two CHD risk factors (154). Bile acid sequestrants have been used as drugs of first choice because they are not absorbed and side effects and adverse events are unlikely unless very large doses are used. Constipation and interference with vitamin absorption are potential difficulties (160). The need for multiple unpalatable doses makes adherence difficult, and most children will not comply with long-term sequestrant use. Doses up to 12 g should be considered (160). Optimum levels of LDL-C are seldom achieved with sequestrants. The triglyceride level may increase significantly in children with combined hyperlipidemia, as it does in adults; consequently, sequestrants should be used with caution.
Lovastatin is the only statin currently approved for children, but simvastatin, pravastatin, and atorvastatin have also been used. LDL-C decreased by 17% on 10 mg/d and by 27% on 40 mg/d of lovastatin in boys 1014 yr of age (161) without an adverse effect on growth and development. LDL-C was reduced by 41% and apo-ß by 34% with 40 mg/d of simvastatin (162) in 173 boys and girls with heterozygous hypercholesterolemia. Other statin studies in children using simvastatin, lovastatin, and pravastatin have consistently demonstrated efficacy, short-term safety, and tolerability (160). Atorvastatin has been used safely in conjunction with diet and cholestyramine for 3 yr in heterozygous familial hypercholesterolemia boys (162, 163). All reached the appropriate LDL-C treatment goal, and adrenal and gonadal function were not altered.
Combination of colestipol with pravastatin is more effective than colestipol alone in reducing LDL-C in children, but compliance is suboptimal (161). Statins will probably be used with increasing frequency to treat childhood dyslipidemia because of their short-term safety and lack of effect on growth and development (162).
Diabetes mellitus Patients with type 2 diabetes mellitus without a prior history of CHD are at the same risk of a CV event as patients without diabetes who have CHD (164). Prevention of CHD should be based on the assumption that diabetes is a CHD equivalent; therefore, this represents secondary and not primary prevention (10). Aggressive reduction of LDL-C to less than 100 mg/dl should be the goal of therapy. Based on observational studies, it may be worth considering reduction of LDL-C to 7080 mg/dl in patients with diabetes (165). All major statin trials (primary and secondary) have demonstrated that diabetics are as likely to benefit from statin therapy as nondiabetics, although the absolute risk reduction is greater than in nondiabetics. This leads to a smaller number needed to prevent an event and greater cost efficacy. Most, but not all, of the statin trials included diabetics, but the numbers varied and were relatively small except for the HPS (42). Also, except for the HPS, the trials were not specifically designed to evaluate CHD events in diabetics. In the HPS, 6000 diabetics, 4000 without CHD and 2000 with CHD, were enrolled. Diabetics were as likely to benefit as nondiabetics across all LDL-C levels (42). In two of the three trials included in the Prospective Pravastatin Pooling Project (PPP), diabetics with LDL-C below 125 mg/dl were at increased CHD risk and benefitted from therapy with pravastatin (166). The HPS and PPP emphasize the importance of treating near optimal levels of LDL-C in diabetics, even in the presence of dyslipidemia. The 627 diabetics enrolled in the VA-HIT also benefitted from gemfibrozil therapy (92). Average baseline lipids in this study were: LDL-C, 110 mg/dl; TG, 150 mg/dl; and HDL-C, 32 mg/dl. These are similar to the values of diabetic patients in CARE that benefitted from pravastatin. Fenofibrate reduced progression of coronary atherosclerosis and clinical events (underpowered for this endpoint) in approximately 400 diabetics (93). Thus, it seems to us that treatment to goal or lower with statin therapy is the first priority in patients with diabetes and addition of a fibrate or niacin should subsequently be based on non-HDL-C. It may be likely, because of insulin resistance and the propensity to dLDL, that LDL-C and triglyceride should be reduced to lower levels than nondiabetics. Most patients with type 2 diabetes have had insulin resistance and the metabolic syndrome for years, and it should not be surprising that treatment of lipid/lipoprotein abnormalities is more effective than glycemic control for preventing CHD events. For a variety of reasons, it is also crucial that the best possible glycemic control be obtained.
Renal insufficiency Patients with renal insufficiency and ESRD are at increased CHD risk. Lipid/lipoprotein abnormalities are common and no doubt contribute to this risk. It has always been assumed that patients with renal insufficiency/failure would benefit from hypolipidemic therapy, but most patients had suboptimum changes in lipids/lipoproteins, and it was not clear whether CHD events would be reduced. A post hoc analysis of 1711 patients in the CARE trial with chronic renal insufficiency (creatinine clearance, <75 ml/min), half of whom received 40 mg pravastatin/d and the other half placebo, revealed a 28% reduction in major CHD events and a 35% reduction in the need for revascularization (167). Benefits were independent of the severity of renal insufficiency, and the therapy was well tolerated.
HIV Insulin resistance, lipodystrophy, and dyslipidemia are common in patients with HIV being treated with protease inhibitors (168, 169). Because HIV is now a controllable chronic disease, the potential long-term adverse CHD effects of these metabolic complications have increasing importance. It is reasonable to believe that such patients are or will be at increased CHD risk (170, 171), the recent study not withstanding (172). Risk factors should be identified and treated (173). Application of ATP III guidelines are reasonable and prudent (173). HIV patients who are dyslipidemic should be treated, although experience with drug regimens, drug combinations, and efficacy is small (174, 175). It is our opinion that HIV therapy complicates but does not prohibit therapy of important lipid/lipoprotein abnormalities and may be very important if protease inhibitors promote atherosclerosis independent of metabolic changes (176).
Summary of Key Points
![]() |
Footnotes |
---|
Received March 6, 2003.
Accepted March 12, 2003.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|