Better safe than sorry


KK Sulik, Bowles Center for Alcohol Studies, University of North Carolina, Chapel Hill, NC, USA 27599-7178


Please cite this paper as: Sulik K, O’Leary-Moore S, Riley E. Better safe than sorry. BJOG 2012;119:1159–1161.

This commentary is written in response to five articles in this volume that report the results of analysis of the impact of binge and low to moderate maternal alcohol consumption on measures of intelligence, attention and executive function. For each of the reports the same group of women and their 5-year-old children from the Danish National Birth Cohort and the same general study design have been used. Although one of the studies found a significant association between maternal consumption of nine or more drinks per week and a heightened risk of lower overall attention scores, the authors generally conclude that small to moderate amounts of alcohol consumed occasionally may not present a serious concern. It is important to note that the conclusions drawn are based on the sample population under study, not on individuals.

Commendably, in these articles it is stated that ‘acceptable levels of alcohol use during pregnancy have not been established and conservative advice for women continues to be to avoid alcohol use during pregnancy’. Indeed with the difficulties inherent in human epidemiological studies with respect to reliable and honest recall of alcohol consumption levels and patterns, knowledge of date of conception, generalisability of data across ethnic and racial groups (i.e. genetic factors) and confounding environmental factors, it remains highly unlikely that for any individual pregnancy a known safe amount of maternal alcohol consumption will be identified. This is the premise upon which the US Surgeon General’s warning states that women who are pregnant or who are considering pregnancy should abstain from alcohol use.

It is well established that individuals adversely affected by prenatal alcohol exposure show a range of structural and functional brain abnormalities, all of which are included as part of fetal alcohol spectrum disorder (FASD). Much of our current understanding of these alcohol-induced abnormalities is summarised in a number of articles comprising a recent special volume of Neuropsychology Review.1 In FASD, functional alterations include deficiencies in cognition that entail not only global reductions in measures of intelligence (e.g. IQ), but also decrements in executive function, motor skills, verbal and non-verbal memory, attention, visual–spatial skills and language. These deficits are related to underlying alterations in regional brain volume, cortical thickness aberrations, fibre tract changes and differences in patterns of neural activation as measured by functional magnetic resonance imaging. In addition, secondary disabilities and psychopathology among individuals with FASD are known and include psychiatric and mental health disorders, social impairments and elevated rates of substance use and abuse.2 Other adverse consequences of gestational alcohol exposure include ocular defects, hearing deficits, early fetal loss, growth retardation and epilepsy.3 Sudden infant death is also linked to maternal alcohol use.4

Studies employing animal models significantly add to our knowledge of the type and extent of damage that prenatal alcohol exposure can cause. Importantly, animal-based investigations can be controlled in ways that are impossible for human studies. While dosage, pattern and developmental stage at exposure are key factors contributing to ethanol teratogenesis, other important controllable variables include genotype and environmental influences including maternal nutrition and stress. With regard to nutrition, vitamin A, folate, choline and compounds with antioxidant properties are all implicated in rodent studies as being capable of diminishing the teratogenic effects of alcohol, supporting the premise that micronutrient deficiency is one factor that may exacerbate alcohol teratogenicity.5 Regarding stress, non-human primate studies have been particularly informative. Notably, examination of adolescent rhesus monkeys prenatally exposed to moderate levels of alcohol, psychological stress, or both alcohol and stress, showed that the combination of prenatal stress and alcohol yielded the most devastating behavioural consequences.6 Although behavioural studies in animals have provided valuable information, the ability of these studies to show changes in specific elements of cognitive functioning that can be reliably extrapolated to humans remains less than ideal. This makes definition of a threshold dosage, if one exists, most unlikely.

Advances in small animal imaging capabilities are furthering the utility of model systems for increasing our understanding of alcohol-induced structural and functional damage.7 An example is the application of high-resolution magnetic resonance imaging and diffusion tensor imaging to a mouse FASD model. Examining narrow windows of exposure, this work provides clear evidence that maternal alcohol treatment limited to periods of development that occur before the time when most human pregnancies are recognised (i.e. in the 3rd week after fertilisation in humans) yields a wide spectrum of brain damage. Abnormalities resulting from this very early alcohol insult include basal and cortical forebrain grey and white matter deficiencies, with the corpus callosum being among the affected fibre tracts. Importantly, there appears to be a very high degree of concordance between the mouse and human magnetic resonance imaging studies7 (Figure 1). Also in keeping with the animal-based findings are the results of a recent study of a Norwegian population sample showing that maternal binge drinking (defined as five drinks per occasion) during weeks 0–4 after conception is predictive of abnormal behaviour.8 These results strongly support the advice to avoid binge drinking when planning pregnancy.

Figure 1.

 Colour-coded diffusion tensor images from mice and magnetic resonance images of human brains illustrate prenatal alcohol exposure-induced dysmorphology. Compared with a control adolescent mouse (A), the brains of two age-matched mice whose mothers had been treated acutely with alcohol at a time corresponding to the third week after conception in humans (C, E) illustrate varying degrees of dysmorphology. Reductions in the corpus callosum (white arrows), particularly involving its median aspect (star in C), are apparent in both alcohol-exposed animals. In the more severely affected individual (E) only the anterior and posterior portions of the corpus callosum remain. The hippocampal commissure (yellow arrows) is also reduced in the more severely affected mouse (E). The corpus callosum (black arrows in B, D, F) is also reduced in the individuals with fetal alcohol syndrome (D, F) compared with normal (B). The dysmorphology in human and mouse brains appears to be comparable. Human images courtesy of Dr S. Mattson, (figure reproduced with permission from O'Leary-Moore et al., Neuropsychol Rev, © Springer Publishing.7).

In the face of existing human and animal data, it is alarming that a large percentage of pregnant women report drinking alcohol.9 Also alarming is the fact that binge drinking among women of childbearing age is prevalent and apparently increasing around the globe.10 Given that approximately 50% of pregnancies are unplanned, it should probably not be surprising that recent studies in the USA11 and Europe12 suggest that FASD occurs in as many as 2–5% of school children.

With alcohol being the human drug of choice, at least in societies that are aware of the potential of alcohol to cause prenatal damage, the desire for a ‘safe’ amount to be identified is pervasive. Is it one drink, two drinks, seven or nine? The fact that the definition of a standard drink in countries like the USA (14 g), Denmark (12 g), the UK (8 g) and Japan (19.75 g) is widely variable (, along with other potential individual risk factors addressed above, makes finding a universally correct answer impossible. Undoubtedly, it is better for women at all stages of pregnancy to be safe by avoiding alcohol consumption rather than being sorry for inadvertently damaging their children.

Disclosure of interests

The authors have no potential conflicts of interest to declare.