Epidemiologic and experimental studies in drinking and pregnancy: the state of the art.

Epidemiologic and experimental studies of the consequences of maternal drinking during pregnancy confirm reports that alcohol in high doses is teratogenic. Alcohol consumed in lower doses has been correlated with intrauterine growth retardation, increased risk of anomalies, behavioral decrements in infants and children, and increased risk of fetal wastage. While evidence linking these effects to moderate levels of alcohol use is growing, there is little agreement on the actual dose at which risks begin, due to the questionable validity of self-reported consumption. Many of the subtle effects correlated with lower amounts of alcohol--growth retardation, minor anomalies, and behavioral alterations--are the same effects that occur in severe degree with high amounts. Thus a dose-response curve may be emerging, although its existence must be substantiated by further studies.