Article Text

Download PDFPDF

Low dose aspirin lowered stroke risk but not risks of MI or cardiovascular deaths in women
  1. John W Eikelboom, MD
  1. McMaster University, Hamilton, Ontario, Canada

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

 Q Is low dose aspirin effective for the primary prevention of cardiovascular disease in women?

    Clinical impact ratings GP/FP/Primary care ★★★★★★☆ IM/Ambulatory care ★★★★★☆☆ Neurology ★★★★★★☆ Haematology ★★★★★★☆


    Embedded ImageDesign:

    randomised, placebo controlled trial (Women’s Health Study).

    Embedded ImageAllocation:


    Embedded ImageBlinding:

    blinded {healthcare providers, participants, data collectors, and outcome assessors}.*

    Embedded ImageFollow up period:

    mean 10 years.

    Embedded ImageSetting:

    USA and Puerto Rico.

    Embedded ImageParticipants:

    39 876 women ⩾45 years of age (mean age 55 y) who had no history of coronary artery disease, cerebrovascular disease, cancer (except non-melanoma skin cancer), or other major chronic illness; contraindication to the study medications; were not taking aspirin, non-steroidal anti-inflammatory drugs, anticoagulants, or corticosteroids; and were not taking vitamin A or E, or ß carotene supplements more than once per week.

    Embedded ImageIntervention:

    aspirin, 100 mg every other day (n = 19 934) or placebo (n = 19 942).

    Embedded ImageOutcomes:

    first major cardiovascular event (non-fatal myocardial infarction, non-fatal stroke, or cardiovascular death); individual cardiovascular endpoints; and adverse events.

    Embedded ImagePatient follow up:

    97% (intention to treat analysis).


    The table shows the results.

    Aspirin v placebo for primary prevention of cardiovascular disease in women at mean 10 years*


    Low dose aspirin lowered risk of stroke but not risks of myocardial infarction or death from cardiovascular causes.

    Abstract and commentary also appear in ACP Journal Club.


    Uncertainty has existed on whether aspirin or warfarin should be the preferred antithrombotic in selected patients at high risk for non-cardioembolic stroke, and whether aspirin is effective and safe for primary prevention of cardiovascular disease (CVD) in women.

    The WASID trial did not show any difference in efficacy between warfarin (target INR 2.0–3.0) and aspirin for preventing stroke or non-stroke vascular death in patients with symptomatic major intracranial artery stenosis. However, the study had to be stopped early because of evidence of harm with warfarin.

    The high incidence of bleeding in patients treated with warfarin (5/100 patient y) is surprising. This may be due to the broad definition of major bleeding. The high dose of aspirin used in this study (1300 mg/d) may have selectively increased bleeding in patients treated with aspirin,1 reducing the difference between the 2 treatments. Nevertheless, a clear excess of bleeding with warfarin remained evident. The increased all cause mortality with warfarin compared with aspirin is unexplained. Previous randomised trials suggest that moderate intensity warfarin (INR 2.0–3.0) is at least as effective as aspirin for preventing major vascular events.2 Yet, warfarin in the WASID trial was associated with a consistent pattern of excess deaths from both vascular and non-vascular causes, including myocardial infarction (MI), sudden death, and cancer. This is unlikely to be explained by the efficacy of high dose aspirin because, unlike safety, no convincing evidence exists to suggest that aspirin’s efficacy is dose related.1,3 The optimal dose of aspirin is less clear from these data but, when considered in the context of what we already know about aspirin, a strong argument can be made to use the lowest proven effective dose (75–150 mg/d), thereby minimising the risk of bleeding complications.

    In 2002, the US Preventive Services Task Force concluded that good evidence existed to suggest that aspirin lowers the incidence of coronary artery disease in adults without previous symptomatic CVD.4 This conclusion was based on a review of 5 randomised trials involving >50 000 persons. However, aspirin did not reduce stroke and the evidence was less certain for women because only 20% of trial participants were women.

    The Women’s Health Study randomised almost 40 000 initially healthy women and, in contrast to previous trials, aspirin did not reduce MI or death but lowered the incidence for ischaemic stroke and transient ischaemic attack. These benefits were partly counterbalanced by an increase in gastrointestinal ulcers and bleeding, highlighting the potential for toxicity even with low doses of alternate daily aspirin, and the importance of balancing risks and benefits when making decisions about the use of aspirin for primary prevention of CVD.

    Controversy concerning apparent sex differences in the antiplatelet effects of aspirin is not new. Subgroup analyses from early randomised aspirin trials suggested that men, but not women, benefited from aspirin;5,6 later, large trials and systematic reviews3 confirmed a benefit in both. In the Women’s Health Study, the low event rates and use of a potentially suboptimal aspirin dose may have contributed to the apparent lack of benefit of aspirin for preventing MI. However, lack of evidence of benefit is not the same as evidence of lack of benefit; the 95% CI of the risk estimates do not exclude a 16% reduction in MI or a 20% reduction in major CVD with aspirin treatment. Nevertheless, if a benefit exists, it is small in absolute terms.


    View Abstract


    • * See glossary.

    • Information provided by author.

    • For correspondence: Dr J E Buring, Brigham and Women’s Hospital, Boston, MA, USA.

    • Sources of funding: National Heart, Lung, and Blood Institute and National Cancer Institute.

    Linked Articles

    • Glossary
      BMJ Publishing Group Ltd