Importance of Complete Abstinence from Alcohol during Pregnancy: Enough Evidence for Justification?

Matthew Hicks; Suzanne Tough


Expert Rev of Obstet Gynecol. 2009;4(4):401-414. 

In This Article

Abstract and Introduction


The association between daily and binge alcohol-drinking episodes during pregnancy and the range of malformations and deficits seen in fetal alcohol spectrum disorder is well described in the literature. However, there is little evidence for low-dose prenatal ethanol exposure and child deficit. This article uses a structured approach to consider the evidence for an association between low-dose prenatal alcohol exposure and neurodevelopmental deficits in children. There is consistent evidence in numerous animal experiments for lasting CNS changes associated with low levels of exposure to alcohol. In epidemiological studies, there is evidence of differences in behavior in children exposed to low doses of alcohol. The results of animal and human studies support current recommendations for complete abstinence from alcohol after conception.


In 1968, Lemoine et al. carefully described a group of children in France born to mothers with histories of alcohol abuse.[1] They noted several distinguishing features, including craniofacial abnormalities, growth restriction and neurocognitive deficits. A few years later, Jones and Smith coined the term fetal alcohol syndrome (FAS) to describe a similar group of children in Seattle (WA, USA).[2] More recently, the term fetal alcohol spectrum disorders (FASDs), a description rather than a diagnosis, has been used to describe a range of deficits that can accompany prenatal alcohol exposure.[3,4] The disabilities include diagnoses captured by FASDs, for example, FAS, partial FAS, alcohol-related birth defects and alcohol-related neurodevelopmental disorders.[3,4] There are also deficits in motor and mental development associated with prenatal alcohol exposure that do not meet the diagnostic criteria for a FASD. These are believed to represent the most common nongenetic cause of mental, learning and behavioral disabilities in North America and are serious lifelong conditions.[5–8,201] In as much as any health issue related to behavior is preventable, these deficits are theoretically 100% preventable.

The reported conservative estimates of the prevalence of FAS/FASD for urban North American populations is 0.23 to three cases per 1000 live births for FAS, and approximately 9.1 per 1000 live births for FASD.[9–14] The reported prevalence in rural communities, foster-care systems and juvenile justice systems can be much higher with rates of 7.2–233 per 1000 found in high-risk populations in rural Canada, while rates of 39.3–98.0 per 1000 live births have been reported in rural South Africa.[15–19] Most recently, a case-finding study in Italian urban schools found a prevalence for FAS of 3.7–7.4 per 1000 children and a prevalence of FASD of 20.3–40.5 per 1000 children.[20]

The impact of FASD is wide reaching, touching the life of the individual and the lives of family members and society as a whole, with major economic, social and medical impact.[6,21,202] The estimated additional lifetime costs associated with education, disability, assisted living, incarceration and healthcare per individual with FAS or a FASD vary by method of calculation and jurisdiction between CAD$1 million and 3.0 million.[10,22,202] Researchers most recently estimated the overall burden on the Canadian economy for FASD at CAD$5.3 billion annually.[23] A case-finding study in the juvenile justice system found that, at a minimum, 23.3% of youth remanded for forensic psychiatric assessment in a 1-year period had a diagnosis of a FASD.[19]

Fetal alcohol spectrum disorders result from alcohol use during the prenatal period. During the first trimester, sometimes before pregnancy recognition, 12–60% of women report alcohol use.[6,24] A proportion of women, estimated to be between 4 and 27%, continue to drink alcohol during the remainder of pregnancy, and the effects on the fetus can be devastating.[6,25–30,201,203] In Canada, Tough et al. described a population-based sample of Canadian women of childbearing age, in which 80% consumed alcohol before conception. Half of all women continued to drink after conception but before they recognized that they were pregnant, and approximately 20% had a binge-drinking episode during this period.[31] Thus, some Canadian women maintain an alcohol-use pattern until pregnancy recognition that potentially places their fetuses at risk for deficits associated with prenatal alcohol exposure. Furthermore, women over 30 years of age were more likely to continue drinking small amounts of alcohol following pregnancy recognition.[31] Similar patterns of alcohol consumption before and during pregnancy have been seen in the UK.[32]

The association between frequent binge drinking and consistent 'high' levels of alcohol consumption during pregnancy with mental, motor and behavioral delays is well established.[33,34] However, it is unclear if and how much of a deficit occurs with doses of one drink per day or occasional drinking early and throughout pregnancy. Recent systematic reviews by Henderson et al. of human observational studies of the effects of low-to-moderate and binge-drinking prenatal alcohol exposure found no consistently significant effects on a range of outcomes, including birth outcomes and birth defects.[35,36] The authors did conclude that there was a possible effect of binge exposure on neurodevelopment.[35,36] Methodological weaknesses were noted in a majority of the studies reviewed, leading the authors to conclude that these weaknesses preclude assuming that low-to-moderate alcohol exposure is safe.[35,36] Since no safe level of alcohol consumption during pregnancy has been determined, health and medical organizations in North America currently recommend that women abstain from alcohol if they are pregnant or are attempting to conceive.[37–42,201,204] However, in the UK and Australia, the recommendation allows for some drinking during pregnancy. Thus, there are different messages from governments and researchers in the Western world ( Table 1 ).

Adding to the confusion, a recent paper by Kelly et al. that was reported extensively in the lay media, found that mothers who reported one to two drinks of alcohol per week or per occasion had children with higher cognitive ability scores and lower behavioral difficulties scores (i.e., less problematic behavior) at 3 years of age when compared with the children of abstainers.[43] What are clinicians, researchers and health policy makers to make of this, let alone lay-people who are reading the press? Is the current recommendation of complete abstinence appropriate? Furthermore, is there any evidence that would suggest that the recommendations of the UK and Australian governments expose fetuses to risk of deficit? Clearly, the answer to these questions could indicate if there is a public health issue of great magnitude, involving many at-risk pregnancies. The purpose of this paper is to consider the evidence regarding the association between low-to-moderate alcohol use during pregnancy and developmental deficits in children. In part, this article uses the structured and thoughtful approach first proposed by Sir Austin Bradford Hill to determine if there is enough evidence to support current recommendations of complete abstinence during pregnancy.[44] Several factors will be considered, including specificity, biological plausibility and coherence, dose–response, experimental evidence and temporality. This review will conclude with specific recommendations based on the evidence and a discussion of areas for future research.

At the outset, it is important to briefly discuss the challenges in assessing the literature in this area. A key and overarching issue in human studies related to prenatal alcohol exposure is in quantifying or assigning an alcohol dose. A major methodical issue in the majority of studies performed to date is that alcohol exposure is averaged across a pregnancy or trimester. This eliminates the ability of researchers to distinguish the impact of binge exposure to alcohol from a low daily alcohol exposure. For example, a weekly average dose of one drink per day does not differentiate between seven drinks consumed on one occasion in a week and one drink consumed every day of the week. This is an important distinction as there is some evidence that a binge exposure may place a fetus at greater risk of deficit. In addition, some studies assign exposure descriptors of 'low', 'moderate' and 'high' dose. However, without evidence for a threshold for deleterious impact on the developing fetus, these descriptors are meaningless. Given that there are limited studies that have examined low-dose alcohol, it is important to acknowledge that this review includes studies of higher exposures and FAS to allow for the extrapolation of potential relationships and expected trends.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.