Despite the stringent criteria for the selection of studies to be included, the following caveats should be noted: Some samples included in the analysis may not be entirely random because participants were screened for certain medical disorders (e.g., diabetes, heart conditions, and stroke) that have been related to alcohol in addition to other factors.Samples with allele distributions that do not meet Hardy-Weinberg equilibrium4 (which are marked in the table summarizing the results) should be viewed with caution because the genotype distribution in these studies is not consistent with the expected distribution for a general sample.
