The review included 12 studies (n=12,238): 5 cohort studies (n=5,416), 2 nested case-control studies (n=4,034) and 5 retrospective case control studies (n=2,788).
The authors found little evidence of publication bias from the funnel plots and statistical tests.
Aspirin was associated with a reduced risk of prostate cancer. The pooled OR for total incidence of prostate cancer was 0.85 for prospective studies (95% confidence interval, CI: 0.77, 0.94) and 1.01 for retrospective studies (95% CI: 0.86, 1.18).
The protective effect of aspirin was stronger for advanced cancers (OR 0.7, 95% CI: 0.52, 0.94; heterogeneity P=0.967) than for total incidence of prostate cancer (OR 0.9, 95% CI: 0.82, 0.99; heterogeneity P=0.317). However, the effects varied by geographic region and study design.
Studies of mixtures of different NSAIDs had less consistent findings: the odds ratio was 0.87 (95% CI: 0.61, 1.23) for non-aspirin NSAIDs (heterogeneity P=0.005) and 0.67 (95% CI: 0.37, 1.22) for mixture of aspirin and non-aspirin NSAIDs (heterogeneity P<0.001).
The authors reported further details of subgroup analyses to account for heterogeneity between the studies. There was no consistent information available about the effect of dose and duration of NSAIDs.