In conclusion, we propose the ADH1B variant rs1229984 as a candidate genetic instrument to study the offspring effects of maternal alcohol drinking, in European origin populations. The association of this SNP with alcohol has been established in several adult populations (23,24). Here and for the first time, we present evidence that this effect exists also just before and during pregnancy. However, replication of the present findings in an independent sample is warranted for the ultimate validation of the proposed instrument. A comparable approach to ‘instrument identification’ has been adopted successfully by Freathy and colleagues, showing that a nicotine receptor gene variant predicts smoking cessation during pregnancy (64).
