Sixth, no formal corrections were made for multiple testing. However, each of the variables chosen in our analysis has strong prior evidence of an association with alcohol consumption. The pattern of findings is not consistent with stochastic effects, although it is always possible that a few of our significant results arose from chance.
