Author + information
- Published online March 3, 2004.
- Expert Panel/Writing Group*,
- Lori Mosca, MD, PhD, (Chair)†,
- Lawrence J Appel, MD†,
- Emelia J Benjamin, MD†,
- Kathy Berra, MSN, ANP†,§,
- Nisha Chandra-Strobos, MD†,
- Rosalind P Fabunmi, PhD†,
- Deborah Grady, MD, MPH¶,
- Constance K Haan, MD∥∥,
- Sharonne N Hayes, MD‡,
- Debra R Judelson, MD#,
- Nora L Keenan, PhD††,
- Patrick McBride, MD, MPH†,
- Suzanne Oparil, MD†,
- Pamela Ouyang, MD†,
- Mehmet C Oz, MD†,
- Michael E Mendelsohn, MD†,
- Richard C Pasternak, MD†,
- Vivian W Pinn, MD§§,
- Rose Marie Robertson, MD†,
- Karin Schenck-Gustafsson, MD, PhD†,
- Cathy A Sila, MD†,
- Sidney C Smith Jr, MD¶¶,
- George Sopko, MD, MPH‡‡,
- Anne L Taylor, MD**,
- Brian W Walsh, MD∥,
- Nanette K Wenger, MD† and
- Christine L Williams, MD, MPH†
In addition, endorsed by: American Academy of Physician Assistants; American Association for Clinical Chemistry; American Association of Cardiovascular and Pulmonary Rehabilitation; American Diabetes Association; American Geriatrics Society; American Society for Preventive Cardiology; American Society of Echocardiography; American Society of Nuclear Cardiology; Association of Women’s Health, Obstetric and Neonatal Nurses; Canadian Women’s Health Network; Jacobs Institute for Women’s Health; Black Women’s Health Imperative; National Women’s Health Resource Center; The North American Menopause Society; Partnership for Gender-Specific Medicine; Preventive Cardiovascular Nurses Association; Sister to Sister: Everyone Has a Heart Foundation, Inc.; Society for Women’s Health Research; Society of Geriatric Cardiology; The Mended Hearts Inc; WomenHeart the National Coalition for Women With Heart Disease; and Women’s Health Research Center.
Significant advances in our knowledge about interventions to prevent cardiovascular disease (CVD) have occurred since publication of the first female-specific recommendations for preventive cardiology in 1999 (1). Despite research-based gains in the treatment of CVD, it remains the leading killer of women in the United States and in most developed areas of the world (2,3). In the United States alone, more than one half million women die of CVD each year, exceeding the number of deaths in men and the next 7 causes of death in women combined. This translates into approximately 1 death every minute (2). Coronary heart disease (CHD) accounts for the majority of CVD deaths in women, disproportionately afflicts racial and ethnic minorities, and is a prime target for prevention (1,2). Because CHD is often fatal, and because nearly two thirds of women who die suddenly have no previously recognized symptoms, it is essential to prevent CHD (2). Other forms of atherosclerotic/thrombotic CVD, such as cerebrovascular disease and peripheral arterial disease, are critically important in women. Strategies known to reduce the burden of CHD may have substantial benefits for the prevention of noncoronary atherosclerosis, although they have been studied less extensively in some of these settings.
In the wake of the reports of the Women’s Health Initiative and the Heart and Estrogen/Progestin Replacement Study (HERS), which unexpectedly showed that combination hormone therapy was associated with adverse CVD effects, there is a heightened need to critically review and document strategies to prevent CVD in women (4–7). These studies underscore the importance of evidence-based practice for chronic disease prevention. Optimal translation and implementation of science to improve preventive care should include a rigorous process of evaluation and clear communication about the quantity and quality of evidence used to support clinical recommendations. Recently, there has been an increase in the number and proportion of women that have participated in clinical trials, although many early CVD prevention trials did not fully include women and other important subpopulations (8). Therefore, it is important to consider the full range of available evidence, including data on men as appropriate, to develop recommendations for diverse populations of women. Furthermore, because many patients seen in clinical practice may have characteristics that are not similar to those of clinical trial participants, it is necessary to draw inferences about the likelihood that data will generalize from research to clinical settings.
The objective of this collaborative effort was to develop the first set of evidence-based guidelines for the prevention of CVD in adult women with a broad range of cardiovascular risk. The technology for identifying CVD in its earliest stages has improved over the past decade, and this has led to a blurring of the distinction between primary and secondary prevention. The concept of CVD as a categorical, “have-or-have-not” condition has been replaced with a growing appreciation for the existence of a continuum of CVD risk. Table 1illustrates a spectrum of CVD, showing risk groups defined by their absolute probability of having a coronary event in 10 years according to the Framingham Risk Score for women (9)Clinical diagnoses and scenarios that broadly group women into categories of high, intermediate, and lower risk also are provided. This scheme allows healthcare providers to match the intensity of risk intervention to the baseline level of CVD risk. A scoring sheet for use in clinical practice to calculate absolute 10-year CHD risk in women is provided in Appendix I. The recommendations herein are designed to assist healthcare providers in optimizing CVD preventive care for all women age 20 years and older. Implementation of these guidelines may differ among countries and regions for cultural, medical, and economic reasons. In addition, application of these guidelines should also take into consideration individual factors such as frailty and life expectancy.
Selection of expert panel members
The leadership of each of the 13 American Heart Association (AHA) Scientific Councils was asked to nominate a recognized expert in CVD prevention who had particular knowledge about women. The president of the AHA appointed at-large members to fill gaps in specific areas of expertise. The AHA Manuscript Oversight Committee approved the chair of the Expert Panel. On the basis of recommendations of the AHA Expert Panel, major professional or government organizations with a mission consistent with CVD prevention were solicited to serve as cosponsors and were asked to nominate 1 representative with full voting rights to serve on the Expert Panel. Panelists also suggested diverse professional and community organizations to endorse the final document after its approval by the AHA Science Advisory Coordinating Committee and cosponsoring organizations.
Selection of topics and candidate recommendations
The Expert Panel reviewed previously published AHA recommendations for the primary and secondary prevention of CVD and discussed and debated topics that were timely, with the goal of developing a set of candidate recommendations for searching and rating (1,10–11). A list of preselected recommendations was circulated to the panel, and experts were asked to independently rate the priority of the recommendation and suggest modifications to the wording. Recommendations were then selected for the systematic literature search.
Systematic search and summary of data
Inclusion and exclusion criteria for studies to be evaluated as part of the evidence-rating process were established according to the Expert Panel recommendation to focus on major CVD clinical end points (death, myocardial infarction, stroke, revascularization procedure, congestive heart failure, or a composite CVD end point) in high-quality studies. The importance of other outcomes, such as quality of life and resource utilization, was recognized, but these were not feasible to include in this version. The purpose of the clinical recommendations is to provide guidance with regard to risk-reducing interventions; therefore, the panel supported the inclusion of studies that were interventional rather than etiologic in nature. For example, studies of the impact of weight loss on major clinical CVD outcomes were included but not studies that simply related obesity to CVD. Inclusion criteria were randomized clinical trials or large prospective cohort studies (>1000 subjects) with CVD risk-reducing interventions evaluated. Also, meta-analyses that used a quantitative systematic review process were included. All studies had to have at least 10 cases of major clinical CVD end points reported. Studies with surrogate end points were excluded unless they met the minimum number of outcome events. Studies meeting the above criteria were included whether or not there were female participants.
The systematic search was conducted by the Duke Center for Clinical Health Policy Research, Durham, NC. Search terms were constructed for each clinical recommendation, with an “explode” term to include related articles. Three databases were searched electronically on OVID, including Medline (1966 through July 3, 2003), the Cumulative Index to Nursing & Allied Health (CINAHL) (1982 through July 3, 2003), and PsycInfo (1872 through July 3, 2003). More than 99% of the studies were located in Medline. Nearly 7000 titles and abstracts identified through the systematic search were reviewed to exclude those that did not meet obvious eligibility criteria or were not available in English. More than 1200 articles were obtained for full-text screening and reviewed for inclusion/exclusion criteria. A standardized abstraction form was completed to document the study design, end points, and decision to include or exclude. Table 2lists the number of articles included/excluded for each category of recommendation.
Included articles were abstracted for more detailed information on a standardized form that included study type, number of participants (% female) at baseline, population characteristics (primary prevention, secondary prevention, or mixed), mean age (age range), percentage diabetic, percentage white, intervention(s) (for drug trials, information was listed about dose, schedule, and duration), primary outcomes including numbers of events, subgroup analysis of clinical end points in women (if analysis available), and comments about important methodological or quality issues.
Expert Panel members reviewed the summary evidence tables for completeness. Tables were updated with publications that were inadvertently omitted or included during the systematic search to comprise the final evidence tables. In addition, results of trials or meta-analyses published subsequent to the systematic search that met inclusion criteria were made available to the Expert Panel. A complete listing of references reviewed by the Expert Panel and used to compile the evidence summary tables is listed in Appendix II. The evidence summary tables are located in an online-only Data Supplement at http://www.circulationaha.org.
Evidence rating system
Two primary reviewers from the Expert Panel were assigned to each candidate recommendation to propose an initial evidence rating and suggest modifications to wording on the basis of the results of the systematic evidence search. A series of conference calls was held to discuss the rating and revised wording of recommendations. Each expert received a final copy of the evidence tables and voted independently on the strength of the recommendation (Class I, IIa, IIb, or III) and level of evidence (A, B, or C) as outlined in Table 3. Class I interventions should be administered unless contraindicated. Class III interventions should not be administered for CVD prevention. The rationale for the rating system is based on methods used in AHA/American College of Cardiology clinical practice guidelines as described (12). The experts also evaluated the likelihood that data from men would generalize to women with regard to each specific risk-reducing intervention (1, very likely; 2, somewhat likely; 3, unlikely; and 0, unable to project). Criteria to determine generalizability were based on factors such as differences in the epidemiology and pathophysiology of CVD between men and women (eg, the ratio of hemorrhagic stroke to coronary events may alter the risk-to-benefit ratio of aspirin in primary prevention for women versus men). The final rating of evidence was determined by a majority vote.
Evidence-based recommendations for the prevention of CVD in women are listed in Table 4. Each recommendation is accompanied by the strength of recommendation, level of evidence to support it, and the generalizability index. The strength of the recommendation is based on not only the level of evidence to support a clinical recommendation, but also on factors such as feasibility of conducting randomized controlled trials in women. Recommendations are grouped in the following categories: lifestyle interventions; major risk factor interventions; atrial fibrillation/stroke prevention; preventive drug interventions; and a Class III category, where routine intervention for CVD prevention is not recommended.
Several lifestyle interventions were rated as Class I recommendations, although the supporting evidence was in many cases classified as level B. These decisions reflect the availability of observational studies as evidence to support the recommendation, as well as ethical issues that preclude conducting randomized controlled trials of certain lifestyle interventions. For example, the Expert Panel regarded smoking cessation as a top priority in clinical practice and suggested that the absence of trial data should not preclude a strong emphasis on clinician interventions to help women stop smoking. More detailed information on how to treat tobacco dependence is available at http://www.surgeongeneral.gov/tobacco/treating_tobacco_use.pdf(Table 5).
Lifestyle interventions received Class I recommendations from the panel not only because of their potential to reduce clinical CVD, but also because heart-healthy lifestyles may prevent the development of major risk factors for CVD (13). Prevention of the development of risk factors through a positive lifestyle approach may minimize the need for more intensive intervention in the future.
Although evidence to support a clinical benefit for CVD event reduction was limited with some interventions (eg, treatment of depression), there may be other important benefits associated with these therapies that are reflected in the strength of the recommendation, such as improved quality of life. Behavioral interventions may have benefits that are not captured by our stringent outcome criteria for clinical CVD events. Weight management via lifestyle and behavioral approaches was rated as a Class I recommendation, level B. The panel suggested there was insufficient evidence to rate more aggressive medical and surgical approaches that generally are limited to a small subset of women.
Our dietary recommendations emphasize intake of a variety of heart-healthy foods. The panel concluded that intake of fish has been associated with a reduced risk of CVD. The benefits of fish seem to result, at least in part, from omega-3 fatty acids. Nonetheless, women of childbearing age, especially pregnant women, should avoid shark, swordfish, king mackerel, and tilefish because the relatively high content of mercury in these fish may impair fetal neurological development. Still, these women can eat other kinds of fish, such as catfish, flounder, and salmon, which have less mercury. For a more complete listing of mercury levels in different types of fish, see the US Food and Drug Administration web site at http://www.cfsan.fda.gov/~frf/sea-mehg.html(Table 5). Women who do not eat fish might consider nonmarine sources of omega-3 fatty acids, such as flaxseed oil, walnut oil, canola oil, soybean oil, or walnuts. However, there is less evidence supporting a cardiovascular benefit from these sources of omega-3 fatty acids (14).
Other expert panels and organizations (including the National Cholesterol Education Program Adult Treatment Panel III [NCEP ATP III]; the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [JNC 7], and the American Diabetes Association) have addressed control of major risk factors extensively and can be referred to for more specific information about management approaches (Table 5)(9,15,16). For example, our recommendation to encourage an optimal blood pressure through lifestyle approaches should be implemented using more detailed information from the JNC 7 report about weight management, adopting a DASH (Dietary Approaches to Stop Hypertension) eating plan, dietary sodium reduction, physical activity, and moderation of alcohol consumption (15). Similarly, NCEP ATP III provides algorithms for cholesterol management and is updated as new evidence becomes available (9). According to NCEP/ATP III, LDL cholesterol is the primary target of lipid-lowering therapy, and intensity of therapy should be matched to the absolute risk of the patient. Glycemic control received a Class I recommendation from the Expert Panel. Treatment of hyperglycemia has been shown to reduce or delay complications of diabetes such as retinopathy, nephropathy, and neuropathy, which underscores the importance of glycemic control in diabetic patients (16). Moreover, both lifestyle intervention and (to a lesser degree) metformin therapy have been shown to reduce the incidence of diabetes (17).
Although there was good consensus on the use of aspirin (75 to 162 mg) in high-risk women, recommendations for aspirin therapy in intermediate- and lower-risk women were more challenging. The difficulty in developing these recommendations was due to the lack of data from primary prevention trials that included women and the possibility that data on men may not necessarily be extrapolated to women. Uncontrolled hypertension is not uncommon in women, and aspirin therapy may increase the risk of hemorrhagic stroke in this setting. Moreover, the risk of gastrointestinal bleeding and other side effects may outweigh the potential benefits of aspirin in women at lower risk for CVD. The panel suggested a conservative approach, pending the results of ongoing clinical trials. It was also noted that nonsteroidal antiinflammatory medications should not be substituted for aspirin for CVD prevention. For stroke prevention among women with atrial fibrillation, a dose of 325 mg of aspirin is needed if there is a contraindication to warfarin therapy or if the risk of a stroke is considered low (<1% annual event rate per year). Tools to determine stroke risk are available at http://www.nhlbi.nih.gov/about/framingham/stroke.htm(Table 5).
The Class III recommendations on hormone therapy and antioxidant supplementation were based on recent clinical trials showing no benefit for CVD prevention and possible adverse effects of these interventions. The panel acknowledged that major trials have been limited to specific types and dosages of these agents, and those results may not generalize to compounds not tested in clinical studies. In particular, ongoing trials will give more information about unopposed estrogen therapy and clinical outcomes. However, given the unproven benefit and possible harm associated with postmenopausal hormone therapies, it was suggested that a conservative approach be taken in clinical practice unless further research is available to support use for CVD prevention. The use of hormone therapy for menopausal symptoms has been addressed by other professional societies (18,19). Although hormone therapy is not recommended for CVD prevention, women and their healthcare providers should weigh the potential risks of therapy against the potential benefits for menopausal symptom control.
The process of developing clinical guidelines has several limitations, even when a systematic approach is undertaken. Most importantly, data used to establish recommendations might be generated from populations that do not reflect the characteristics of the patient being treated, and individual responses can vary significantly. The clinical cardiovascular end points chosen for inclusion in the systematic evaluation do not necessarily reflect the net clinical impact and do not include many end points that are clinically important but often not reported (eg, symptoms, quality of life, functional status, hospitalizations, resource utilization, etc). We simplified the recommendation for each level of risk for purposes of clinical utility and acknowledge that there might be variability in efficacy and effectiveness of various interventions within the same risk intervention category (eg, various doses or types of physical activity or drugs within the same class may yield different results). The Framingham risk score may not apply equally to all populations, but it performs well within subgroups (20,21). We may have omitted or included some studies because of the limitations of electronic searching and human error; however, the likelihood that such an inadvertent omission or inclusion would alter a recommendation is small. Our recommendations are based on evidence available to the panel through November 2003, and as science evolves, recommendations may have to be revised. Finally, we do not include a comprehensive plan for implementation of the guidelines in this document. The AHA is developing professional education programs and other initiatives to facilitate the dissemination and implementation of the guidelines.
Conclusions and future directions
Overwhelming evidence suggests that CVD can be prevented in both women and men. Clinical recommendations are provided to assist healthcare providers and the public in efforts to avoid an initial or recurrent cardiovascular event. Strategies to implement these guidelines and prioritize risk-reducing therapies in clinical practice are outlined in Tables 6 and 7. ⇓⇓Our systematic search of the literature shows that several prevention strategies are likely to have substantially greater benefit than risk and that some interventions are likely to be associated with greater risk than benefit. It is important that the public be appropriately informed about potentially lifesaving preventive therapies and take action to lower their risk. On the basis of our review of the scientific evidence, it appears the risk of no action is far greater than that of applying knowledge to prevent CVD. Approximately 75% of the original research articles that met our inclusion criteria included female subjects, and very few presented race/ethnic-specific analyses. Moreover, few studies included elderly women, especially those over 80, in whom CVD is common. The results of this project highlight the need to include diverse populations in research studies and to present subgroup analyses so that guidance can be tailored, if appropriate, to subpopulations. These recommendations are meant to assist clinicians on the basis of our current state of evidence and supersede previous AHA prevention guidelines with regard to women (1,10,11,22). Because health care is a blend of science and art, we emphasize that guidelines are not a substitute for good clinical judgment.
We are grateful to the Foundation for the Advancement of Cardiac Therapies (FACT) Foundation, Palm Beach, Fla, for providing financial support for the literature searches and to the Duke Center for Clinical Health Policy Research, Durham, NC, for conducting and summarizing the systematic searches. Persons from Duke who contributed to this project include Rowena J. Dolor, MD, MHS; L. Kristin Newby, MD, MHS; Lori A. Bastian, MD, MPH; Mike Blazing, MD; Ann J. Brown, MD, FACE; Tracy W. Gaudet, MD; Richard S. Liebowitz, MD; Kenneth W. Mahaffey, MD; Alison A. Lee, MA, MPH; Rebecca Gray, DPhil; Ayn C. Huntington, BA; Allison Meyer; Mimi Sengupta Biswas, MD, MHSc; Daniel R. Bensimhon, MD; Adi Cohen, MD; Suneet Kaur, MD; Abhi Goyal, MD; Svati H. Shah, MD; John L. Petersen, MD; Jonathan E.E. Yager, MD; Jean-Pierre Dery, MD, FRCP(C); and Camille G. Frazier, MD. The assistance of Lisa Rehm in coordinating this manuscript is greatly appreciated.
↵1 The American Heart Association makes every effort to avoid any actual or potential conflicts of interest that may arise as a result of an outside relationship or a personal, professional, or business interest of a member of the writing panel. Specifically, all members of the writing group are required to complete and submit a Disclosure Questionnaire showing all such relationships that might be perceived as real or potential conflicts of interest.
This statement was approved by the American Heart Association Science Advisory and Coordinating Committee and by the American College of Cardiology Foundation in December 2003. This document is available on the World Wide Web sites of the American College of Cardiology (http://www.acc.org) and the American Heart Association (http://www.americanheart.org). A single reprint is available by calling 800-242-8721 (US only) or writing the American Heart Association, Public Information, 7272 Greenville Ave, Dallas, TX 75231-4596. Ask for reprint No. 71–0276. To purchase additional reprints: up to 999 copies, call 800-611-6083 (US only) or fax 413-665-2671; 1000 or more copies, call 410-528-4121, fax 410-528-4264, or e-mail . To make photocopies for personal or educational use, call the Copyright Clearance Center, 978-750-8400.
(J Am Coll Cardiol 2004;43:900–21)
The online-only Data Supplement is available athttp://www.acc.org.
- American College of Cardiology Foundation and the American Heart Association, Inc.
- Mosca L,
- Grundy S.M,
- Judelson D,
- et al.
- American Heart Association
- Yusuf S,
- Reddy S,
- Ounpuu S,
- et al.
- Rossouw J.E,
- Anderson G.L,
- Prentice R.L,
- et al.
- Smith S,
- Blair S,
- Bonow R,
- et al.
- Pearson T.A,
- Blair S.N,
- Daniels S.R,
- et al.
- Gibbons R,
- Smith S,
- Antman E
- Benjamin E.J,
- Smith S.C,
- Cooper R.S,
- et al.
- Kris-Etherton P.M,
- Harris W.S,
- Appel L.J
- American College of Obstetricians and Gynecologists
- Grundy S.M,
- D’Agostino R.B Sr.,
- Mosca L
- Mosca L,
- Collins P,
- Herrington D.M,
- et al.