3 Reasons Why Scientists Can’t Seem to Agree About Alcohol and Pregnancy

By
Photo: Richard Drew

The question of alcohol and pregnancy is a divisive one, to put it delicately. On the one hand — no one is seriously arguing that heavy or binge drinking is a safe or healthy thing to do during pregnancy. But what about light or moderate drinking? It gets murkier here. Some studies suggest that a couple drinks a week are probably harmless; others have argued that as many as eight per week are fine. And yet the nation’s top health officials firmly disagree, recommending that women who are pregnant should abstain from alcohol. If scientists can’t agree on this, perhaps it’s not so surprising that the resulting policies can be so confusing — like last week’s reminder from the New York City Commission on Human Rights that, um, yeah, pregnant women are allowed in bars.

The question of drinking while pregnant is mired in so much social stigma, which can vary by culture. But the research itself is also part of the issue here. A new book, The Informed Parent — written by science reporters Tara Haelle and Emily Willingham — dives deep into the best existing research on pregnancy and early childhood, and a rather large section of the book is devoted to the back-and-forth data about drinking during pregnancy. As it turns out, there are some fairly concerning methodology issues that pervade this type of research. Here’s a brief overview:

Too many studies conflate women who always abstain from alcohol with women who only abstain during pregnancy. That’s a bigger deal than it may initially appear to be, because there are some important differences between these two groups. In one of the relatively rare studies to separate all-time abstainers from pregnancy-only abstainers, researchers in Denmark tracked about 63,000 mothers, and found that the women who typically drank at least a little — but stopped during pregnancy — tended to be a little healthier than their teetotaling peers. The all-time abstainers, on the other hand, were less likely to have healthy habits like regularly exercising; they were also more likely to have some not-great habits, like watching more television and drinking more soda. This group as a whole was also more likely than the pregnancy-only abstainers to have diabetes or asthma, and they also were more likely to be overweight.

But perhaps most important, these women were more likely to have a mental-health condition, something that could certainly be affecting the findings of reports that do not separate the all-the-time abstainers from the pregnancy-only abstainers. “It’s easy to see how teetotalers’ higher rate of psychological problems — a known risk factor for behavioral problems in children — might cancel out the ability to detect any possible behavioral effects among the children of light drinkers in pregnancy, especially when light drinkers typically have more education, higher incomes, healthier weights, and more fish on the dinner table,” Haelle and Willingham write.

Most of the research relies on self-reporting. Okay, this is not strictly a problem with this one area of research — it’s also a problem in practically all corners of medical and psychological studies. Self-reporting is just what it sounds like: Researchers who collect data this way depend on their study participants to be honest and accurate when answering questions about their health or mood or behavior, and oftentimes volunteers are … not, whether because of a simple lapse in memory or because they’re perhaps trying to preserve their own self-image.

As Haelle and Willingham point out, self-reporting is pretty much the only way to track the information needed in these studies on pregnancy and alcohol, since “alcohol intake is usually self-reported because no biomarkers are available to track ethanol intake,” they write. “Self-report involves limitations, including concern about stigma, knowledge of how much ethanol an alcoholic beverage contains, or how much was in a drink (mixed drinks, for example, can be quite variable in ethanol content), and when the mothers were asked (during or after pregnancy).”

Most studies don’t track kids for long enough. Even many of the best-conducted studies out there only follow the children of mothers who drank — or didn’t — during pregnancy through early childhood. This is no good, Haelle and Willingham note, because some key signs of behavioral problems don’t show up until late childhood. One study they found, for example, that just ten percent of 5-year-old children with fetal alcohol syndrome showed signs of attention problems — but by their 10th birthday, 60 percent of them had attention issues.

And, really, these are just a few of the confounding factors making this so hard to study. There is also the tiny fact that it would be wildly unethical to conduct a randomized controlled trial to study the effects drinking alcohol while pregnant. (A randomized controlled trial, or RCT, is generally considered the best type of study scientists can conduct, and it involves randomly assigning study participants to one condition or another.) And there is also the matter of genetics to contend with: It’s possible that some women, and some babies, are more affected by the potentially damaging effects of alcohol than others.

Haelle and Willingham refrain from giving advice or recommendations, choosing to stick to the research, but they do close with this: “Ultimately, if there is a ‘safe’ lower threshold of how much alcohol a pregnant woman can drink and have absolutely no effect on her embryo or fetus, we may never know what it is.”

The Issues With Research on Drinking and Pregnancy