Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: 6 clinical conditions can be modified to reduce risk for a first stroke

ACP J Club. 1999 Sept-Oct;131:30-31. doi:10.7326/ACPJC-1999-131-2-030

Source Citation

Gorelick PB, Sacco RL, Smith DB, et al. Prevention of a first stroke. A review of guidelines and a multidisciplinary consensus statement from the National Stroke Association. JAMA. 1999 Mar 24/31;281:1112-20. [PubMed ID: 99202546]



What evidence supports clinical recommendations for prevention of a first stroke in primary care patients?

Data sources

English-language studies were identified with MEDLINE (1990 to November 1998) by using the terms guideline; consensus; cerebrovascular disorders; and risk factors plus primary prevention for cerebrovascular disorders, hypercholesterolemia, and hyperlipidemia. Guidelines and consensus documents from 6 journals (Stroke, Hypertension, Circulation, Diabetes Care, Diabetes, and Neurology) were reviewed. Bibliographies from studies, systematic review articles, guidelines, textbooks, and reference guides were reviewed; other nonjournal publications and Internet sites were scanned.

Study selection

Randomized controlled trials (RCTs) and meta-analyses were assessed. 6 risk factors for a first stroke were evaluated: hypertension; coronary artery disease, including blood lipid levels; atrial fibrillation (AF); diabetes mellitus; asymptomatic carotid artery stenosis; and lifestyle variables (cigarette smoking, alcohol consumption, physical activity, and diet).

Data extraction

Clinical experts in neurology, cardiology, family practice, nursing, physician assistant practices, and health services research extracted data on quality of evidence, patient sample, and first stroke. Evidence and guideline recommendations were assessed, and recommendations were maintained or updated on the basis of new evidence. Members of the expert panel reached a consensus.

Main results

Hypertension. 14 RCTs were identified and analyzed; systematic review by using data from these trials showed that a 5- to 6-mm Hg decrease in diastolic blood pressure reduced the risk for stroke by 42%. 1 RCT showed that decreasing isolated systolic hypertension in elderly persons reduced the risk for stroke by 36%. Systematic review of data from these trials showed that therapy with diuretics (odds ratio [OR] 0.61, 95% CI 0.51 to 0.72) or β-blockers (OR 0.75, CI 0.57 to 0.98) reduced risk for stroke in older persons with hypertension. Recommendations included offering treatment to persons who are most likely to develop stroke (e.g., African-American and elderly persons), checking blood pressure at all clinic visits, and monitoring blood pressure at home for persons with hypertension.

Myocardial infarction. The rate of ischemic stroke after myocardial infarction (MI) is 1% to 2% per year; the greatest risk is in the first month after MI (31%). Several meta-analyses have shown that aspirin therapy after MI reduces risk for nonfatal stroke, although these analyses vary. A meta-analysis of 3 RCTs has shown that warfarin given after MI to achieve an international normalized ratio of 2.0 to 3.0 reduced risk for stroke. Meta-analysis showed that antiplatelet agents in patients with previous MI reduce nonfatal stroke by 39%, nonfatal MI by 31%, and vascular death by 15%. The combined end point of MI, stroke, and vascular death was reduced with a risk difference of 3.2%. 3 RCTs have shown that statins reduce the risk for stroke in patients with previously high lipid levels. Recommendations for patients who have had MI are warfarin for those who also have AF, left ventricular thrombus, or severe left ventricular dysfunction; aspirin for those with no complications; and statins for those with normal to high lipid levels.

NonvalvularAF. AF increases risk for stroke by a factor of 6. Meta-analysis has shown that patients with AF have a 68% (CI 50% to 79%) reduction in the rate of stroke with warfarin use and a 21% reduction with aspirin use. These reductions must be balanced with an increased risk for serious bleeding (1.3%/y for warfarin and 1.0%/y for aspirin). Recommendations for patients with AF include the use of warfarin for those who are older than 75 years of age and those who are 65 to 75 years of age with specific risk factors for stroke. Patients younger than 65 years of age without risk factors should be treated with aspirin.

Diabetes mellitus. Patients with diabetes have an increased risk for stroke. 2 RCTs have shown that tight glucose control in patients with type 1 diabetes and intensive drug use in patients with type 2 diabetes reduce microvascular complications but not macrovascular complications, such as stroke. Tight control of hypertension and type 2 diabetes reduced the risk for fatal and nonfatal stroke by 44%. Recommendations emphasize tight blood glucose control for prevention of microvascular complications in patients with diabetes; more research is needed on prevention of first stroke.

Asymptomatic carotid artery disease. Risk for stroke increases with increasing carotid stenosis. 1 RCT showed that patients who had stenosis of 60% to 99% had a decreased risk for stroke or death after carotid endarterectomy of 5.9% over 5 years. Existing guidelines do not uniformly support endarterectomy for asymptomatic stenosis. Carotid endarterectomy should be considered if the asymptomatic stenosis is ≥ 60%, but only if the site-specific combined surgical morbidity and mortality rate is < 3%.

Lifestyle factors. A meta-analysis of 32 studies showed that cigarette smoking is associated with an increased risk for stroke (relative risk 1.5, CI 1.4 to 1.6). Risk is proportional to the number of cigarettes smoked. Light or moderate alcohol consumption may protect against ischemic stroke, but heavy alcohol use increases the risk for hemorrhagic stroke. Regular exercise reduces the risk for premature death and cardiovascular disease. The protective effect of exercise may be mediated by other risk factors. Some dietary factors may be associated with stroke; for example, increased sodium intake, elevated homocysteine levels, and vitamin deficiency may increase risk, and fruit and vegetable consumption may decrease risk. Recommendations include promotion of smoking cessation, regular exercise, moderate alcohol use, and a healthy diet.


Several interventions reduce the incidence of a first stroke by modifying risk factors.These are the treatment of hypertension, use of aspirin after uncomplicated myocardial infarction, use of warfarin for patients with atrial fibrillation after myocardial infarction, use of statins to control lipid levels after any myocardial infarction, use of warfarin or aspirin for patients with atrial fibrillation (depending on other risk factors), and possible carotid endarterectomy for patients with asymptomatic stenosis ≥ 60%. Observational studies support modification of lifestyle-related risk factors (smoking, exercise, alcohol consumption, and diet).

Sources of funding: Bristol-Myers Squibb Co.; Boehringer Ingelheim Pharmaceutical, Inc.; Sanofi Winthrop Pharmaceuticals.

For correspondence: Dr. P.B. Gorelick, Center for Stroke Research, Rush Medical College, 1645 West Jackson Boulevard, Suite 400, Chicago, IL 60612, USA. FAX 312-432-0937.


Clinical practice guidelines and consensus statements should summarize the best available evidence and provide busy clinicians with practical, readily available, unbiased recommendations about patient care. This is not an easy task because consensus often involves compromise and simplification of complex issues; biases may exist throughout the process. Fortunately, an evidence-based approach to assessing clinical practice guidelines (1) has been developed to determine guideline validity, sensibility, and usefulness in patient care.

Stroke is a major public health issue; therefore, the development of this primary prevention consensus statement by the National Stroke Association (NSA) is timely and relevant. The methods section is complete, and the consensus process is well described. Unfortunately, no explicit information (except financial disclosures) was provided about the experts who evaluated the evidence and their levels of training. This may have introduced bias in the use of the data. In addition, no mention was made of an author with expertise in statistics.

An exhaustive, clear search strategy for relevant evidence is important to the success of a review. The description of the NSA strategy is ambiguous and limited and could have led to incomplete results. For example, keyword could mean either "MeSH (Medical Subject Heading) heading" or "textword", and consensus is not a MeSH term. Bias may have been introduced by restricting the search to English-only guidelines listed in MEDLINE. It may also have been appropriate to search the Cochrane Library.

It is difficult to identify and include all available evidence or relevant ongoing research. 2 examples of material not included in this review are the Medical Research Council and British Heart Foundation Heart Protection Study (an ongoing RCT of cholesterol reduction with simvastatin, antioxidant vitamin therapy, or both in 20 000 participants at high risk for coronary artery disease [2]) and a recent systematic review of 5 RCTs of endarterectomy for asymptomatic stenosis that included 2440 patients (3) (1659 patients were included in the Asymptomatic Carotid Atherosclerosis Study done in North America [4]). Review of the 5 RCTs showed that carotid surgery reduced the risk for ipsilateral stroke or death by 38%; 50 operations were needed to prevent 1 additional ipsilateral stroke or death over 3 years. Although the risk for perioperative stroke or death from carotid surgery seems to be lower for persons with asymptomatic stenosis (2.3%) than for persons with symptomatic stenosis, the risk for stroke without surgery is low, and for most persons, the balance of risk and benefit from surgery is unclear.

The decision to base this consensus statement on guidelines derived from RCTs and systematic reviews is sound. However, no criteria for quality assessment of the guidelines, the underlying RCTs, or systematic reviews were given, and we are not told how the data were combined in the final document. A systematic approach to synthesis of the evidence is important and should be explicit. It is also difficult to critically compare the various guidelines because insufficient data are presented. The meta-analysis by the Antiplatelet Trialists Collaboration (5), which suggested a significant benefit from aspirin use, was based on 8 RCTs (16 747 patients and 211 outcomes). The less favorable overview by the Matchar and colleagues (6) evaluated 3 RCTs (9600 patients and 111 outcomes). It seems reasonable that aspirin, although not as effective as warfarin in the prevention of stroke after MI, is still worth considering in some patients.

Despite these methodologic shortcomings, the paper assesses 6 clinically important, potentially modifiable risk factors and, for the most part, considers clinically reasonable treatments and alternatives. Even without strong evidence from RCTs, few would argue with such sensible recommendations as smoking cessation, moderation of alcohol intake, regular exercise, and ensuring a healthy diet.

Gordon Gubitz, MD
Western General HospitalEdinburgh, Scotland, UK


1. Hayward RS, Wilson MC, Tunis SR, Bass EB, Guyatt G. Users' guides to the medical literature. VIII. How to use clinical practice guidelines. A. Are the recommendations valid? The Evidence-Based Medicine Working Group. JAMA. 1995;274: 570-4.

2. MRC/BHF Heart Protection Study [Abstract]. Stroke. 1994;25:543.

3. Benavente O, Moher D, Pham B. Carotid endarterectomy for asymptomatic carotid stenosis: a meta-analysis. BMJ. 1998;317:1477-80.

4. Executive Committee for the Asymptomatic Carotid Atherosclerosis (ACAS) Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA.1995;273:1421-8.

5. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet therapy. I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ. 1994;308:81-106.

6. Matchar DB, McCrory DC, Barnett HJ, Feussner JR. Medical treatment for stroke prevention. Ann Intern Med. 1994;121:41-53.