Skip navigation

Aspirin and stroke

Because of its antiplatelet actions, aspirin is now well established in preventing subsequent myocardial infarction and ischaemic stroke in patients with a range of cardiovascular conditions. Some studies have suggested that it may increase the risk of haemorrhagic stroke, and so a meta-analysis of studies has been done [1] to try and measure this risk.


The authors performed extensive searching for randomised studies of aspirin versus control and which had stroke as an outcome. To be included studies had to have:


They found 16 trials with over 55,000 subjects. The mean aspirin dose was 273 mg/day (range 75 - 1500 mg/day) and the mean duration of treatment was 37 months (1 to 72 months). The study was predominantly in white men (88% men, 99% white) with a a mean age of 59 years.

The 16 studies differed in size and the rate at which haemorrhagic stroke was found (Figure). Overall the rate was 0.26% with aspirin and 0.12% with control. This translates into a number needed to harm with aspirin to cause a haemorrhagic stroke of 715 (95% CI 471 to 1483).

But the rate of ischaemic stroke was lower (1.68%) with aspirin than with control (2.14%). The number needed to treat with aspirin to prevent an ischaemic stroke was 217 (145 to 428).
Effect of aspirin on haemorrhagic, ischaemic and total stroke
Stroke subtypes Stroke with aspirin (%) Stroke with control (%) Absolute risk/benefit per 10,000 (95%CI) Number needed to harm or treat (95%CI)
Haemorrhagic 75/28570 (0.26) 33/26892 (0.12) 12 (5 to 20) 715 (471 to 1488)
Ischaemic 480/28570 (1.68) 576/26892 (2.14) -39 (-17 to -61) 217 (145 to 428)
Total stroke 703/28570 (2.46) 742/26892 (2.76) -31 (-5 to -51) 335 (177 to 3043)

For all strokes, including some that were unclassified, the balance was favourable. With aspirin 2.46% of patients had a stroke, compared with 2.76% with control. This is a number needed to treat to prevent any type of stroke of 335 (177 to 3043).


Even in this large analysis of 55,000 people, there were only 108 cases of haemorrhagic stroke. It seems reasonable to conclude that although there is a slight increase in the rate of haemorrhagic stroke in patients treated with aspirin, this is outweighed by the benefit conferred by the reduction in ischaemic strokes, which were far more numerous.


  1. J He, PK Whelton, B Vu, MJ Klag. Aspirin and risk of hemorrhagic stroke. A meta-analysis of randomized controlled trials. JAMA 1998 280: 1930-1935.

previous or next story in this issue