highly significantP < 0.01
However, the 2007 meta-analyses of 4 observational studies reported that coffee consumption could significantly reduce AD risk in comparison with nonconsumers with a pooled risk estimate of 0.73 (95% CI = 0.58–0.92), in a highly significant heterogeneity (chi-squared: 13.6, P < 0.01) [ 29 ]. 2.4.