PURPOSE:

To estimate the pooled risk of coffee consumption for Alzheimer's disease (AD).

MATERIAL AND METHODS:

We have reviewed all observational studies that evaluated the association between AD risk and coffee consumption. Four studies were identified: two case-control studies and two cohorts. These studies were carried out between 1990 and 2002.

RESULTS:

There was an obvious protective effect of coffee consumption in the pooled estimate [risk estimate: 0.73 (95% confidence interval: 0.58-0.92)]. However, the homogeneity test was highly significant (p<0.01), indicating heterogeneity across the pooled studies. Pooled analysis applying the random effect model was 0.79 with 95% confidence interval overlapping unity (95% confidence interval: 0.46-1.36). Three studies assessed coffee consumption by interview questionnaire. The risk of AD in coffee consumers versus non-consumers in studies that used interview questionnaire had a pooled risk estimate of 0.70 with 95% confidence interval 0.55-0.90.

CONCLUSION:

Although our pooled estimates show that coffee consumption is inversely associated with the risk of AD, the four studies had heterogeneous methodologies and results. Further prospective studies evaluating the association between coffee consumption and AD are strongly needed.