Should a healthy older adult take low-dose aspirin to prevent a first heart attack or stroke?

The idea has an intuitive appeal. Aspirin reduces the ability of platelets—small blood components involved in clotting—to stick together. Because a blood clot can block an artery and cause a heart attack or ischaemic stroke, making platelets less active could prevent an event.

The same action also makes bleeding more likely. That trade-off looks very different in someone who has already had a cardiovascular event than it does in someone who has not. Recent long-term Australian evidence helps explain why daily aspirin is no longer treated as a routine healthy-ageing habit.

Primary and secondary prevention are different decisions

Primary prevention means trying to prevent a person’s first cardiovascular event. Secondary prevention means reducing the chance of another event in someone with established cardiovascular disease, such as a previous heart attack or stroke.

That distinction changes the starting risk. After a cardiovascular event, the chance of another is usually high enough that antiplatelet treatment can offer substantial benefit. The 2025 Australian acute coronary syndromes guideline therefore includes antiplatelet therapy in specific regimens after an acute coronary syndrome or coronary intervention, with duration adjusted for the person’s ischaemic and bleeding risks.

Before a first event, the absolute cardiovascular risk is lower. A similar relative effect then prevents fewer events, while aspirin’s bleeding effect remains. Its over-the-counter familiarity does not alter that arithmetic.

What the primary-prevention trials found

A 2019 systematic review combined 13 randomised trials involving 164,225 people without established cardiovascular disease. Aspirin was associated with fewer cardiovascular events: 57.1 versus 61.4 events per 10,000 participant-years. The hazard ratio was 0.89, with a 95% credible interval from 0.84 to 0.95.

The relative reduction sounds larger than the absolute difference. Across the trials, the absolute risk reduction was 0.38%, equivalent to treating about 265 people to prevent one composite cardiovascular event over the study periods.

Major bleeding moved in the other direction: 23.1 versus 16.4 events per 10,000 participant-years. The hazard ratio was 1.43, with a 95% credible interval from 1.30 to 1.56. The absolute increase was 0.47%, equivalent to one additional major bleed for about every 210 people treated.

Those numbers cannot be reduced to a simple exchange of one heart attack for one bleed. The composite cardiovascular outcome included cardiovascular death, non-fatal heart attack and non-fatal stroke; major bleeding definitions also covered serious events in different organs. Their consequences vary. The figures do show why a modest relative cardiovascular effect is not enough to establish net benefit for everyone.

ASPREE tested the question in healthy older adults

The ASPREE trial is particularly relevant in Australia. It enrolled 19,114 community-dwelling adults, predominantly Australians aged 70 and over, who did not have cardiovascular disease, dementia or an independence-limiting disability. Participants received either 100 mg of aspirin daily or a matching placebo for a median of 4.7 years.

Cardiovascular disease occurred at 10.7 events per 1,000 person-years with aspirin and 11.3 with placebo. The hazard ratio was 0.95, with a 95% confidence interval from 0.83 to 1.08. In other words, the trial did not show a statistically reliable cardiovascular reduction.

Major haemorrhage occurred at 8.6 events per 1,000 person-years with aspirin and 6.2 with placebo. That was a 38% relative increase, and the confidence interval excluded no difference. Both upper gastrointestinal bleeding and intracranial bleeding were more frequent in the aspirin group.

An extended analysis followed consenting participants for a further median 4.3 years after the randomised treatment period ended. Across the trial and extension, original assignment to aspirin produced no long-term reduction in major adverse cardiovascular events: hazard ratio 1.04, 95% confidence interval 0.94 to 1.15. Major haemorrhage remained higher: hazard ratio 1.24, 95% confidence interval 1.10 to 1.39.

The extension needs careful reading. Once the trial ended, aspirin use was no longer controlled by random assignment; people could start, stop or change treatment. The follow-up strengthens the finding that no delayed cardiovascular advantage emerged, but it cannot tell an individual what will happen after stopping aspirin.

Why absolute cardiovascular risk comes first

Australia’s current cardiovascular prevention guideline begins with a person’s estimated five-year risk, not one risk factor or a familiar tablet. For people without known cardiovascular disease, the Aus CVD Risk Calculator classifies risk as low below 5%, intermediate from 5% to under 10%, and high at 10% or above. The guideline prioritises blood-pressure and lipid-lowering treatment for people most likely to benefit, alongside risk-factor management.

Bleeding risk also needs its own assessment. Age, previous gastrointestinal bleeding or ulcer disease, uncontrolled blood pressure, kidney or liver disease, alcohol use, and medicines such as anticoagulants or anti-inflammatory drugs can shift the balance. Several factors that raise cardiovascular risk also raise bleeding risk, so a high heart-risk estimate does not automatically settle the aspirin question.

ASPREE also studied a selected group. Participants were well enough to live independently and people with a clear aspirin indication, a contraindication or high bleeding risk were excluded. Its findings apply well to the question of routine aspirin in initially healthy older adults, but not to every patient and not to secondary prevention.

The conversation worth having

For someone already taking aspirin, the most useful question is not simply “Is aspirin good or bad?” It is “What is the indication?” A previous heart attack, stroke, coronary stent or another vascular condition puts the decision in a different evidence category from taking it to stay generally well.

A clinician can then clarify the person’s absolute cardiovascular risk, the chance and likely consequences of bleeding, interacting medicines, and whether other proven risk-reduction measures are being used. Aspirin should not be started or stopped independently on the basis of a population study.

The event being prevented and the person’s starting risk determine the calculation. Before a first cardiovascular event, aspirin’s small average benefit can be matched or exceeded by bleeding. After an event, the calculation may be very different.

Sources

Evidence note: This Insight is general education, not individual medical advice. It does not recommend starting or stopping aspirin. A qualified clinician should interpret the indication, cardiovascular risk, bleeding risk and other medicines in context.