Sensitivity analysis showed that no individual group biased the findings to the extent that it could account for the overall strong association. For instance, the results of AD and alcohol abuse were strongly consistent, regardless of the data set removed, with the P values never > 6×10−51 among the 50 studies in the allelic analysis; under the dominant model the results also showed strong consistency, regardless of the data set removed, with the P value never > 3×10−41 among the 52 studies. The results are shown for the allelic and genotypic analyses in supplementary Tables 3 and 4, respectively.
