Should statins be used in primary prevention?
theheart.org April 12 2012 Michael O'Riordan
Baltimore, MD and San Francisco, CA - Differing opinions on the use of statins in primary prevention make the pages of one of the leading medical journals this week, with the Journal of the American Medical Association (JAMA) the latest in a line of professional and mainstream media outlets getting in on the contentious topic [1,2]. Introduced by the JAMA editors to encourage discussion and debate , the inaugural "dueling viewpoints" kicks off its new series by considering the clinical question of whether or not a healthy 55-year-old male with elevated cholesterol levels should begin taking the lipid-lowering medication.
The two "combatants" in the clinical duel will also be familiar, having previously debated the topic in the pages of the Wall Street Journal, as well as on theheart.org. For Drs Rita Redberg and William Katz (University of San Francisco, California), who argue that healthy men should not take statins, there are other effective means to reduce cardiovascular risk, including dietary changes, weight loss, and increased exercise.
"These strategies are effective in increasing longevity and also result in other positive benefits, including improved mood and sexual function and fewer fractures," they write. "Although these strategies are challenging, prescribing a statin may undermine them. For example, some patients derive a false sense of security that because they are taking a statin they can eat whatever they want and do not have to exercise."
In their counterpoint, Drs Michael Blaha, Khurram Nasir, and Roger Blumenthal (Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Baltimore, MD) agree that the cornerstone of treatment for patients with elevated cholesterol levels will always be diet and exercise but that statins can be a "critical adjunct for those identified to be at increased coronary heart disease risk." The Johns Hopkins physicians argue that there is no logic in waiting for an MI to occur before starting statin therapy and that if clinicians are unsure of the risk of seemingly healthy patients with elevated cholesterol levels, the use of coronary artery calcium (CAC) screening can help.
"The CAC scan is a helpful tool that enables clinicians to direct statin treatment at the disease (coronary atherosclerosis) that they propose to treat and illustrates the concept of risk-based, individualized decision making," write Blaha, Nasir, and Blumenthal. "Statin therapy would not be recommended if a CAC scan revealed a score of 0."
In their viewpoint, they point to data from WOSCOPS and AFCAPS/TexCAPS showing reductions in MI and other coronary events in the primary-prevention setting. However, they argue that the debate over cholesterol therapy needs to be rephrased, because doctors should never treat elevated cholesterol levels in isolation but instead aim to provide therapy to the highest-risk patients most likely to benefit.
For Redberg and Katz, however, the data simply do not support the use of statins in the 55-year male patient with normal blood pressure and no family history of disease but with elevated cholesterol levels. They point to a recent meta-analysis in healthy but high-risk men and women showing no reduction in mortality with statin therapy, as well as a recent Cochrane review showing similar results. Moreover, Redberg and Katz highlight the adverse effects associated with statins, including cognitive defects and diabetes.
"For every 100 patients with elevated cholesterol levels who take statins for five years, a myocardial infarction will be prevented in one or two patients," they write. "Preventing a heart attack is a meaningful outcome. However, by taking statins, one or more patients will develop diabetes and 20% or more will experience disabling symptoms, including muscle weakness, fatigue, and memory loss."
JAMA. 2012;307(14):1532. doi: 10.1001/jama.2012.432
The Debut of Dueling Viewpoints
Medical journals should encourage discussion and debate.1 ,2 For many clinical, public health, and policy issues, there seems to be little agreement and limited scholarly discourse. For example, consider the common clinical question of whether an otherwise healthy middle-aged man with an elevated cholesterol level should begin to take a statin. This question is addressed by Michael J. Blaha and colleagues3 and Rita Redberg and Mitchell Katz4 in a new feature in our Viewpoint section.
The ground rules for the authors are simple. The editors will pose a question to them, which the authors can help refine. The authors must begin their answer with a yes or no response to that question so their opinion is clear. The authors will present a scholarly discussion of the issue and will be expected to "defend" their position with evidence. The authors of these Viewpoints will not review each others' article prior to publication.
In addition, each of the paired Viewpoints will be accompanied by an online poll so JAMA readers can participate in the debate and discussion. This online feature, along with other technologic advances, allows us to harness electronic, video, and audio technologies to better serve our readers. Journals can no longer be print centric; they must reach out to their audience in creative new formats. This online feature for these paired Viewpoints is just one example of how JAMA is using technology to engage readers in journal content, enabling them to help create it, sculpt it, and respond to it.
We also invite JAMA readers to pose questions for "experts" to address in this section in the future. The questions do not necessarily have to relate to a clinical issue. For example, we are considering the following query for our next set of paired Viewpoint articles: do patients care if their physician has financial conflicts of interest?
Decision making in medicine is complex and involves careful integration of evidence, experience, and patient preference.5 Some decisions are dominated by evidence; others by experience; and others, such as end-of-life care, must reflect patient preference. The shifting sands of decision making-for the individual patient, hospital, multicenter specialty group, community health center, or entire health care system-require reflection and discussion.
We hope that by providing thoughtful and scholarly exploration of key questions in these Viewpoint articles, and by promoting reader response and interaction about these questions, JAMA will contribute to the robust discussion and debate around important issues in health care.
We look forward to hearing from you about this new addition to the Viewpoints section.
1. Author Affiliations: Dr Bauchner (firstname.lastname@example.org) is Editor in Chief and Dr Fontanarosa is Executive Editor, JAMA.
JAMA. 2012;307(14):1489-1490. doi: 10.1001/jama.2012.425
1. Author Affiliations: The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Baltimore, Maryland.
Corresponding Author: Roger S. Blumenthal, MD, The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Blalock 524C-Cardiology, 600 N Wolfe St, Baltimore, MD 21287 (email@example.com).
Statin Therapy for Healthy Men Identified as "Increased Risk"
Should a 55-year-old man who is otherwise well, with systolic blood pressure of 110 mm Hg, total cholesterol of 250 mg/dL, and no family history of premature CHD be treated with a statin?-Yes.
Atherosclerotic coronary heart disease (CHD) is the most common cause of morbidity and mortality in the world. The "lipid hypothesis" of CHD is clearly established: (1) circulating cholesterol plays a central role in atherogenesis and is an integral component of the requisite lesion, the coronary plaque; (2) cholesterol levels beginning in childhood predict lifetime risk of atherosclerotic CHD events in a dose-response relationship; and (3) statins lower cholesterol levels and reduce CHD and cerebrovascular events directly proportional to the degree of low-density lipoprotein cholesterol (LDL-C) lowering. As a result, guidelines from around the world support a combined lifestyle and pharmacologic approach to cholesterol lowering directed at patients with elevated CHD risk.
Assuming a high-density lipoprotein cholesterol (HDL-C) level of about 40 mg/dL, the patient in this common clinical scenario would have an "intermediate" 10-year risk for developing CHD (approximately 10%) based on the Framingham Risk Score (FRS). As always, lifestyle change is the first-line therapy. However, if this patient's cholesterol level remains abnormal, despite sustained attempts at lifestyle optimization, statin therapy should be considered with the goal of reducing CHD risk. Current guidelines suggest an LDL-C goal of less than 130 mg/dL with an optional target of less than 100 mg/dL.1 In the shared decision-making process, the clinician should explicitly inform this patient that a statin is likely to reduce the chance of a first CHD event and reduce the chance of stroke and may offer a survival benefit that is likely to become more evident over a lifetime.
Evidence Supporting Primary Prevention
The WOSCOPS enrolled 6595 men aged 45 to 64 years with no previous history of myocardial infarction and a mean (SD) plasma cholesterol level of 272 (23) mg/dL. Treatment with pravastatin, 40 mg, resulted in a 31% reduction in myocardial infarction and CHD-related death (248 vs 174 events and 135 vs 106 deaths for placebo vs pravastatin, respectively).2
Similarly, the AFCAPS/TexCAPS randomized 6605 asymptomatic adults with a mean (SD) LDL-C level of 221 (21) mg/dL and low HDL-C (36  mg/dL) to lovastatin, 20 to 40 mg, vs placebo. Treatment with lovastatin reduced the incidence of first major coronary events by 37% and myocardial infarction by 40% (183 vs 116 events and 95 vs 57 myocardial infarctions for placebo vs lovastatin, respectively).3
The JUPITER trial enrolled 17 802 healthy men and women with so-called normal LDL-C less than 130 mg/dL and elevated high-sensitivity C-reactive protein greater than 2.0 mg/L. Aggressive lowering of LDL-C in those randomized to receive rosuvastatin, 20 mg, reduced the risk of myocardial infarction, stroke, and revascularization by about 44% (251 vs 142 events for placebo vs rosuvastatin) and total mortality by 20% (247 vs 198 events, respectively). The effect of aggressive LDL-C lowering in JUPITER was substantial considering that the baseline median LDL-C was just 108 mg/dL.4 Subanalysis demonstrated the largest absolute reduction in patients with a FRS of 11% to 20% (145 vs 74 events for placebo vs rosuvastatin; hazard ratio [HR], 0.51; 95% CI, 0.39-0.68) followed by those with FRS of 5% to 10% (59 vs 32 events; HR, 0.55; 95% CI, 0.36-0.84).5
Risk-Based, Individualized Treatment Decisions
Nearly all US adults have elevated cholesterol compared to their evolutionary ancestors. The debate over cholesterol therapy must therefore be rephrased. Clinicians should never treat elevated cholesterol levels in isolation. The main goal must be to direct risk-reducing, atherogenic lipoprotein-reducing therapies to those at the highest risk who are more likely to benefit.
What if the patient in this scenario is uncertain about his true risk and thus unclear about the absolute benefit of statin treatment? The best predictor of risk in intermediate-risk patients is the coronary artery calcium (CAC) scan.
Knowing a patient's CAC score, a directly measured marker of the burden of coronary artery disease, enables the clinician to integrate risk exposure over a lifetime and to use this information to guide decision making. High CAC scores (>100) signify higher CHD risk and thus a lower estimated number needed to treat (NNT) with statins. In contrast, a CAC score of 0 equates to very low near-term (5-year) CHD risk and unfavorably high NNT.6 The CAC scan is the single best test for reclassifying intermediate risk patients into their most appropriate treatment groups.
The argument of using a statin for the patient in this scenario can be supported by data from the Multi-Ethnic Study of Atherosclerosis (MESA). A 55-year-old patient with a total cholesterol of approximately 250 mg/dL and a normal blood pressure would have a 50% chance of having a CAC score of 0; this would translate to an estimated 10-year CHD event rate of less than 2% and an estimated 5-year NNT of approximately 300 using a 35% event reduction with statins. However, simple presence of CAC would increase that risk nearly 4-fold and reduce the estimated 5-year NNT to approximately 70. This patient also would have a 13% chance that the CAC score is greater than 100, which would suggest an estimated 10-year CHD event rate greater than 12% and an estimated 5-year NNT of approximately 45.
The CAC scan is a helpful tool that enables clinicians to direct statin treatment at the disease (coronary atherosclerosis) that they propose to treat and illustrates the concept of risk-based, individualized decision making. Statin therapy would not be recommended if a CAC scan revealed a score of 0.
Arguments Against Selective Use of Statins
Some physicians see no role for pharmacologic treatment of elevated cholesterol level to prevent CHD in any asymptomatic patient. What are the main points of contention?
1. Are statins safe? Adverse effects with statin therapy are rare. Approximately 5% of patients will develop muscle-related complaints that are generally reversible after drug discontinuation. Many of these patients can tolerate a different statin. There is no good peer-reviewed evidence that statins lead to cognitive impairment or memory loss, as has been anecdotally reported; one report suggested that statins may improve memory.7 In appropriate middle-aged patients, the risk of type 2 diabetes associated with statins is mainly seen in those with preexisting glucose intolerance and is minimal in comparison with CHD event reduction.
2. Do statins lead to less adherence with a prudent lifestyle? In fact, there is evidence to the contrary; a recommendation from a physician for statin treatment may motivate overall healthy behaviors.8 It is incumbent on physicians to refrain from paternalism/maternalism and to encourage sustained motivation for adherence to both lifestyle and medicine.
3. Is there a durable benefit to statin therapy, or should statins be prescribed only after a myocardial infarction? There is no apparent logic in waiting for a myocardial infarction or a stroke to occur before starting a risk-reducing therapy. A recent meta-analysis of trials confirms that statins retain their benefit after discontinuation of randomized therapy.9
4. Is statin therapy cost-effective? With the emergence of generic high-potency statins like simvastatin (~$4 a month) and atorvastatin, statin therapy is increasingly cost-effective, well below the typical willingness-to-pay threshold. Would it be more cost-effective to spend this money on walking trails, neighborhood renovation, and increased accessibility to fruits and vegetables? This is not likely, despite the critical importance of these approaches.
5. Do statins only work in men? In the recent meta-analysis by Kostis et al,10 women derived just as much benefit from statins as men for primary prevention.
6. Do patients expect medications to prolong survival within 5 years? Most patients do not expect near-term survival benefit from medicine; they are concerned about myocardial infarction, stroke, venous thrombosis, and the resulting chronic disease and disability that may occur. They see their parents, who have vascular dementia and congestive heart failure, and seek safe strategies to reduce their risk. In fact, more than ever, the modern patient is focused on quality of life and not exclusively on longevity.
The cornerstone therapies for patients with elevated cholesterol levels will always be dietary modification and renewed emphasis on physical activity. Statin therapy is a critical adjunct for those identified to be at increased CHD risk.
JAMA. 2012;307(14):1491-1492. doi: 10.1001/jama.2012.423
1. Author Affiliations: Division of Cardiology, Department of Medicine, University of California, San Francisco (Dr Redberg); and Department of Health Services, County of Los Angeles, Los Angeles, California (Dr Katz). Dr Redberg is also Editor, Archives of Internal Medicine. Dr Katz is also Deputy Editor, Archives of Internal Medicine.
Corresponding Author: Rita F. Redberg, MD, Division of Cardiology, University of California, San Francisco, 505 Parnassus Ave, M1180, San Francisco, CA 94143 (firstname.lastname@example.org).
Healthy Men Should Not Take Statins
Should a 55-year-old man who is otherwise well, with systolic blood pressure of 110 mm Hg, total cholesterol of 250 mg/dL, and no family history of premature CHD be treated with a statin? - No.
Extensive epidemiologic data demonstrate that higher cholesterol levels are associated with a greater risk of heart disease. At the population level, higher levels of cholesterol are associated with a diet greater in fatty foods, particularly trans fat and meat, and low intake of fruits and vegetables.
The important questions for clinicians (and for patients) are as follows: (1) does treatment of elevated cholesterol levels with statins in otherwise healthy persons decrease mortality or prevent other serious outcomes? (2) What are the adverse effects associated with statin treatment in healthy persons? (3) Do the potential benefits outweigh the potential risks? The answers to these questions suggest that statin therapy should not be recommended for men with elevated cholesterol who are otherwise healthy.
1. What is the benefit of statin therapy in healthy men with high cholesterol levels? Data from a meta-analysis of 11 trials including 65 229 persons with 244 000 person-years of follow-up in healthy but high-risk men and women showed no reduction in mortality associated with treatment with statins.1 A 2011 Cochrane review of treatment with statins among persons without documented coronary disease came to similar conclusions.2 The Cochrane review also observed that all but one of the clinical trials providing evidence on this issue were sponsored by the pharmaceutical industry. It is well established that industry-sponsored trials are more likely than non-industry-sponsored trials to report favorable results for drug treatment because of biased reporting, biased interpretation, or both of trial results.6
2. What adverse effects are associated with statin treatment in healthy persons? All treatments designed to prevent disease-such as death from coronary disease-can also result in adverse effects. Data from observational studies show much higher rates for statin-associated myopathy and other adverse events in actual use than the 1% to 5% rate reported in clinical trials. This underestimation of adverse events occurs because the trials excluded up to 30% of patients with many common comorbidities, such as those with a history of muscular pains, as well as renal or hepatic insufficiency.3 Many randomized trials also excluded patients who had adverse effects of treatment during an open-label run-in period. For example, in the Treat to New Targets trial, after initial exclusions based on comorbidities, an additional 35% of eligible patients, or 16% of patients, were excluded during an 8-week, open-label, run-in phase because of adverse events, ischemic events, or participants' lipid levels while taking the drug not meeting entry criteria.7 Additionally, the results of randomized trials of statin treatment likely underestimate common symptoms such as myalgia, fatigue, and other minor muscle complaints because these studies often only collect data on more quantifiable adverse effects such as rhabdomyolysis.
· Numerous anecdotal reports as well as a small trial8,9 have suggested that statin therapy causes cognitive impairment, but this adverse outcome would not have been captured in randomized trials. The true extent of cognitive impairment associated with statins remains understudied. It is disappointing that more data are not available on important adverse events associated with statin treatment, despite millions of prescriptions and many years of use. This information could be easily collected in observational studies and from registries. One population-based cohort study in Great Britain of more than 2 million statin users found that statin use was associated with increased risks of moderate or serious liver dysfunction, acute renal failure, moderate or serious myopathy, and cataract.4 The risk of diabetes with statin use has been seen in randomized clinical trials such as JUPITER, which found a 3% risk of developing diabetes in the rosuvastatin group, significantly higher than in the placebo group. In observational data from the Women's Health Initiative, there was an unadjusted 71% increased risk and 48% adjusted increased risk of diabetes in healthy women taking statins.5
3. Do the potential benefits outweigh the potential risks? Based on all current evidence, a healthy man with elevated cholesterol will not live any longer if he takes statins. For every 100 patients with elevated cholesterol levels who take statins for 5 years, a myocardial infarction will be prevented in 1 or 2 patients.7 Preventing a heart attack is a meaningful outcome. However, by taking statins, 1 or more patients will develop diabetes and 20% or more will experience disabling symptoms, including muscle weakness, fatigue, and memory loss.3
Nondrug Approaches to Reducing Coronary Risk
There are effective methods for reducing cardiovascular risk in otherwise healthy men: dietary modification, weight loss, and increased exercise. These strategies are effective in increasing longevity and also result in other positive benefits, including improved mood and sexual function10 and fewer fractures. Although these strategies are challenging, prescribing a statin may undermine them. For example, some patients derive a false sense of security that because they are taking a statin they can eat whatever they want and do not have to exercise.
For some clinicians, evidence that statins reduce the risk of recurrent coronary events in patients with documented coronary disease leads to the belief that statins also "must" be beneficial for patients without coronary disease. However, recent history is rife with examples of interventions that are proven to work in patients with serious disease yet are not efficacious when generalized to patients without serious disease. For example, coronary artery bypass graft (CABG) surgery is lifesaving for patients with symptomatic left main disease. However, CABG surgery would not be a good choice for single-vessel coronary artery disease (CAD) because risks would outweigh benefits in less extensive CAD. Similarly, the benefits of carotid endarterectomy in preventing stroke outweigh the risks for symptomatic patients with tight carotid artery stenosis, but not for asymptomatic patients with less critical stenosis. In addition, the use of aspirin is similar to statins for prevention. The data show clear benefit for aspirin in secondary prevention of cardiovascular disease, but not for primary prevention. Practitioners should not be generalizing from other settings when good data indicate that statins are not effective in improving length or quality of life when used for primary prevention.
For the 55-year-old man in this scenario, his risk of myocardial infarction in the next 10 years based on the Framingham Risk Score varies from 10% to 20%. His risk is driven mostly by his age rather than by his cholesterol level. Increasing age has a much larger influence on risk for cardiovascular disease than do increasing levels of cholesterol. Recent data on increased risk of diabetes, cognitive dysfunction, and muscle pain associated with statins suggest that there is risk with no evidence of benefit. Advising healthy patients to take a drug that does not offer the possibility to feel better or live longer and has significant adverse effects with potential decrement in quality of life is not in their interest.
At the same time, there are significant opportunities for improvement in lifestyle counseling and interventions. Even small changes in diet and increases in physical activity and smoking cessation can lead to significant personal and population health benefits. Such positive lifestyle changes have the key advantage of helping patients feel better and live longer. Lifestyle counseling should remain the focus of primarily prevention efforts-at the physician and public health levels.