Thirty-three studies were eligible, including 1,916,448 individuals (range 55 to 635,031). Nineteen were cohort studies, eight were population-based case–control studies, five were hospital-based case–control studies and one was a RCT.
When all 33 studies were pooled, aspirin was associated with a statistically significant reduction in breast cancer risk (OR 0.86, 95% CI 0.81 to 0.92; Ι²=79%). Results were statistically significant for cohort studies (OR 0.91, 95% CI 0.84 to 0.98; Ι²=81%; 19 studies) population-based case–control studies (OR 0.79, 95% CI 0.68 to 0.90; Ι²=80%; eight studies) and hospital-based case–control studies (OR 0.75, 95% CI 0.65 to 0.84; Ι²=0%; five studies) but not for the single RCT (OR 0.95, 95% CI 0.87 to 1.09).
There was evidence for publication bias when looking at all studies (p=0.02 for Begg rank correlation method).