A systematic review including 8 studies with a total of 31,366 (not all of whom were included in the meta-analyses) subjects was abstracted in DARE. There were 4 prospective studies (n=15,899) and 7 non-prospective studies (n=15,467). The non-prospective studies comprised 3 case-control studies and 4 cross-sectional studies. All of the included studies controlled for some or all of the following potential confounding factors using matching or multivariate statistical analysis: age, female sex, low education, and genetic factors. NSAID exposure was associated with a significantly reduced risk of Alzheimer's disease in both non-prospective studies (OR 0.51, 95% CI: 0.40 to 0.66, P<0.001) and prospective studies (lifetime NSAID use, RR 0.74, 95% CI: 0.62 to 0.89, P<0.01; NSAID use for 2 years or more, RR 0.42, 95% CI: 0.26 to 0.66, P<0.001, based on 3 studies). No significant heterogeneity was found for any of the analyses (P=0.48, P=0.23 and P=0.56, respectively).
Comment: The quality of evidence is upgraded by strong association (for long-term use) and consistency of results. The evidence is only strong enough to show an association and not a causal relationship between NSAID exposure and risk of Alzheimer's disease.