Ten SNPs were selected based on published evidence of functionality (i.e. rs1229984 in ADH1B) or association with alcohol-related phenotypes, and typed as part of an investigation of the role of maternal drinking on childhood growth and development in ALSPAC. Efforts were made to choose a set of independent common polymorphisms (MAF ≥ 0.05 for all except rs1229984). Coverage of ADH1A in particular (3 SNPs), whose related enzymes are most active during fetal and early life (63), reflects the scope of the study for which the genotyping was done. The prior for inclusion in the study for each SNP was not assessed through a formal meta-analysis of published associations, which was impossible because of the limited evidence presented. Rather, we looked for consistency of association across the different studies that reported results for a particular SNP, as well as evidence of functionality. We acknowledge that this strategy has the shortcoming of not being fully replicable.
