Oxford University Press's
Academic Insights for the Thinking World

The consequences of alcohol and pregnancy recommendations

By Sarah CM Roberts and Lyndsay Ammon Avalos


What should be the public health messaging on drinking during pregnancy? The answer isn’t clear-cut. We know that there is strong evidence that high levels of alcohol consumption during pregnancy harm the developing fetus. However, we don’t know conclusively what the impact is of lower level alcohol consumption. That is, we don’t know if there is a truly safe level of alcohol use, nor do we know if the line between safe and unsafe alcohol consumption is the same for all pregnant women.

This inconclusive evidence regarding harms at low levels of drinking presents a challenge for public health officials who are responsible for creating official recommendations regarding alcohol use during pregnancy. What should these messages be, and what level and type of evidence warrants a complete abstinence recommendation? Professionals and official government bodies differ in their recommendations. For example, a recent debate in the British Medical Journal focused on whether it is “all right for women to drink small amounts of alcohol in pregnancy.” The “yes” side argued that, in the absence of conclusive evidence of harms from low-levels of alcohol use, women’s autonomy to make decisions for themselves should be respected. The “no” side argued that abstinence was the safest message. Further, even with access to the same evidence, some countries including the United States recommend complete abstinence during pregnancy; others including the United Kingdom recommend that women not exceed a low-level.

Another ongoing debate in this field centers on the unintended consequences of abstinence recommendations. In particular, some worry that policies recommending abstinence from alcohol during pregnancy will lead women drinking low-levels of alcohol to terminate otherwise wanted pregnancies. On this question, we have some compelling data. Our recent paper, “Alcohol, tobacco, and drug use as reasons for abortion,” examined whether evidence supports this argument. Briefly, in our study of 956 women seeking abortion in the United States, we found 2.5% reported alcohol as a reason for seeking abortion. Almost all (21 out of the 25) of the women reporting alcohol as a reason for were drinking more than a “low” amount. Specifically, their typical alcohol use in the month before they discovered their pregnancies included binge drinking or blacking out from drinking on average once a week. We also found that all the women reporting alcohol as a reason for abortion reported that their pregnancies were unplanned. Thus, in the context of a long-standing policy recommending abstinence from alcohol during pregnancy, we didn’t find evidence that these recommendations lead women using low-levels of alcohol to terminate otherwise wanted pregnancies.

Our findings have already been used to strengthen arguments for abstinence recommendations, and suggest that abstinence recommendations probably do not have the unintended consequence of leading women to unnecessarily terminate otherwise wanted pregnancies. However, to be clear, our findings don’t provide any additional evidence as to whether abstinence or low-level only recommendations actually have intended consequences. Intended consequences may include a decrease in the proportion of pregnant women who drink at all and the proportion of pregnant women who binge drink, drink moderately, or drink heavily. They also may include a reduction in the negative health effects of alcohol use during pregnancy including fetal alcohol syndrome as well as growth restrictions and cognitive impairments among children born to women drinking during pregnancy.

Unfortunately, there hasn’t been much research about whether official recommendations have intended consequences. Two recent Australian studies tried to answer this question. The alcohol and pregnancy recommendations in Australia have changed from abstinence to low-level to abstinence over the past 20 years. Both studies found that more pregnant women drank at all under low-level than abstinence guidelines, although this increase was statistically significant in only one case. Importantly, one of the studies also found that a low-level guideline was associated with a decrease in moderate or heavy drinking during pregnancy.

We need more research to confirm these Australian findings; if they are supported by additional research, we will need to have some difficult conversations. We will need to explicitly discuss whether it is more important to decrease the proportion of pregnant women who drink at all or whether we should concentrate on decreasing the proportion of pregnant women drinking at moderate or heavy levels.

Knowing that abstinence messages don’t seem to lead women to terminate otherwise wanted pregnancies doesn’t tell us what effects the abstinence messages have. We still don’t know whether or how much official abstinence recommendations influence pregnant women’s decisions to drink alcohol at all, or pregnant women’s decisions about how much to drink.  We also don’t know whether abstinence recommendations improve pregnancy outcomes or child well-being. To determine the most appropriate recommendations, we need to move beyond debating the conclusiveness of epidemiologic evidence regarding effects of low-level drinking, searching for a universal safe level, and focusing on unintended consequences of the abstinence recommendation. We need more research about how women interpret and act on official alcohol and pregnancy recommendations, and whether there are demographic or drinking-level-related subgroup differences in interpretation and action. Only when we have these data can we really establish the most appropriate public health recommendations.

Sarah CM Roberts is a public health social scientist at Advancing New Standards in Reproductive Health (ANSIRH) at the University of California, San Francisco, United States. Lyndsay Ammon Avalos is a Research Scientist at Kaiser Permanente Division of Research in Oakland, California, United States. Their recent paper, “Alcohol, tobacco and drug use as reasons for abortion,” has been made freely available for a limited time by the journal Alcohol and Alcoholism.

Alcohol and Alcoholism publishes papers on the biomedical, psychological, and sociological aspects of alcoholism and alcohol research, provided that they make a new and significant contribution to knowledge in the field. It is the official journal of the Medical Council on Alcohol.

Subscribe to the OUPblog via email or RSS.
Subscribe to only health and medicines articles on the OUPblog via email or RSS.

Image credit: Pregnant woman with a glass of wine. Photo by DomenicoGelermo, iStockphoto.

Recent Comments

  1. sudon't

    As we know, all European women consumed alcohol throughout their pregnancies, and indeed, their entire lives, for thousands of years, up until quite recently. People had no access to safe water, nor soft-drinks, nor juices.
    I think it’s safe to call it moderate consumption, in most cases, since we know that wine was mixed with water, and beer-like beverages were brewed with a lower alcohol content, or mixed with water, for daily consumption. I think it’s safe to assume that, for those whose ancestors were long exposed to alcohol, moderate consumption poses no great risk.
    But we must be careful of the message we send, lest people try to think for themselves, and not behave as we wish them to.

  2. [...] here to read this interesting post and book review on The  Consequences of Alcohol and Pregnancy Recommendations. Categories : Fetal Alcohol Spectrum Disorders (FASD) Tags : alcohol and [...]

  3. Anonomous

    The message of abstenince is affecting women. I terminated a pregnancy due to fear of fas. I was low risk. More research needs to be done to protect women. The trauma of termination was unbareable I went from one tragedy of fear of fas to trauma of a termination that I will live with for the rest of my life.

Leave a Comment

Your email address will not be published. Required fields are marked *