The benefit of aspirin for patients with previous symptomatic atherothrombosis of the heart, brain and limb in the secondary prevention of recurrent serious vascular events is well established. However, the role of aspirin in the primary prevention of cardiovascular disease among people who have no symptoms of vascular disease is controversial.1 A recent summary of the evidence has prompted recommendations from the third US Preventive Services Task Force2,3 and the American Heart Association.4

The evidence is based on a systematic review of five randomised controlled clinical trials of the effectiveness and safety of long-term aspirin use (75–500 mg a day or every other day) over 3–7 years in over 50 000 individuals with no previous symptomatic cardiovascular disease.5-9 Most participants were middle-aged men, although there were more than 10 000 women included in two trials and substantial numbers of patients aged 70–80 years in four of the five trials.

Among patients randomly allocated to receive aspirin, the rate of subsequent serious vascular events (non-fatal stroke, non-fatal myocardial infarction, or death due to vascular causes) was reduced significantly, from 4.8% (no aspirin) to 4.2% (aspirin) over about 56 months' follow-up. This is an odds reduction of 13% (95% CI, 5%–19%) and an annual risk reduction of about 0.1% (ie, one serious vascular event avoided per 1000 patients treated with aspirin for one year) (see Box). Most of the benefits of aspirin were due to a significant 28% (95% CI, 13%–40%) reduction in the odds of a coronary event (myocardial infarction or sudden death). There was no reduction in the occurrence of all stroke. However, aspirin was associated with a non-significant 40% (95% CI, −10% to 100%) increase in the odds of haemorrhagic stroke, which is consistent with the excess risk of haemorrhagic stroke seen in secondary prevention trials using aspirin.1 In absolute terms, this represents an excess risk of one haemorrhagic stroke per 10 000 patients treated with aspirin per year.10 Aspirin was also associated with a 70% (95% CI, 40%–110%) increase in the odds of major extracranial (mainly gastrointestinal) haemorrhage, which is an excess of 0.7 (95% CI, 0.4–0.9) major extracranial haemorrhages per 1000 patients treated with aspirin per year (see Box).10

The Preventive Services Task Force concluded that there is now good evidence that aspirin lowers the incidence of coronary heart disease (CHD) in adults who are at increased risk, and that it also increases the incidence of gastrointestinal bleeding.2 It considered that there was fair evidence that aspirin increases the risk of haemorrhagic stroke.2 The evidence was most reliable for men aged 40–75 years and less reliable for women and older men. The optimum dose of aspirin is not known, but dosages of 75–150 mg/day seem to be as effective as higher doses and are associated with a lower risk of adverse gastrointestinal effects.

The American Heart Association recommends low-dose aspirin prophylaxis in people with a 10-year CHD risk of over 10% (ie, > 1% per year),4 whereas the Preventive Services Task Force recommends aspirin for people with a five-year CHD risk of over 3% (ie, > 0.6% per year).2,3 The latter is justifiable, in our view: for every 1000 patients with a 3% risk of a coronary event over five years, long-term aspirin therapy prevented 4–12 coronary events and caused 0–2 haemorrhagic strokes and 2–4 major gastrointestinal bleeding events. This represents a benefit-to-harm ratio of about 2.0 (see Box). The benefit-to-harm ratio of aspirin was most favourable among people at high risk of a future cardiovascular event and low risk of haemorrhagic complications.

The implication for clinicians is that decisions to prescribe aspirin therapy for the primary prevention of cardiovascular events should be based on an assessment of the patient's absolute risk of a vascular event without aspirin, the absolute risk of a gastrointestinal or intracranial haemorrhage with aspirin, and the patient's preference. In addition, decisions about aspirin therapy should be reviewed at least every five years, or when new vascular risk factors are detected. Risk stratification should incorporate specific information about multiple risk factors, rather than simply counting the number of risk factors.11,12

Risk factors for cardiovascular disease include increasing age, being male, cigarette smoking, increasing blood pressure, increasing blood total cholesterol level, decreasing high-density lipoprotein cholesterol level, raised fasting blood glucose level (ie, diabetes mellitus), and a positive family history of cardiovascular disease (in younger adults).4,13

Risk factors for haemorrhagic complications of aspirin include increasing age, any bleeding diathesis, uncontrolled hypertension, and concomitant use of other nonsteroidal anti-inflammatory agents or anticoagulants. Enteric-coated or buffered preparations of aspirin do not clearly reduce adverse gastrointestinal effects.

People at increased cardiovascular risk who may wish to consider long-term aspirin therapy (75–150 mg/day) are men over 40 years of age, postmenopausal women, and younger people with risk factors for cardiovascular disease (eg, hypertension, diabetes).4,13 However, there is still insufficient information to reliably identify the minority of individuals who will benefit and the minority who will be harmed by regular treatment with aspirin. Further information will soon be available from several studies: the Women's Health Study (comparing aspirin 100 mg taken every alternate day with placebo among 40 000 healthy postmenopausal women); the Aspirin in Asymptomatic Atherosclerosis trial (comparing low-dose aspirin with placebo in 3300 middle-aged participants with asymptomatic peripheral atherosclerosis, identified by an ankle brachial pressure index of ≤ 0.9); and the CHARISMA (Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management and Avoidance) trial (comparing aspirin with a clopidogrel/aspirin combination for preventing serious vascular events among about 15 000 people at high risk of cardiovascular disease who are currently taking aspirin).

In the meantime, the challenge for clinicians is to translate the evidence into practice by ascertaining the absolute risk of subsequent serious vascular events for all people who may be at risk, and to prescribe long-term aspirin 75–150 mg/day, with long-term follow-up, for those with an absolute risk of CHD exceeding 3% over the next five years. The challenge for academic clinicians is to devise more valid "risk calculators" that incorporate the risk of serious cardiovascular events and the risk of adverse events, such as intracranial and gastrointestinal haemorrhage.