Emily Oster just came out with a book, Expecting Better, and one of the chapters is about this very topic.
According to her data, if you limit yourself to a glass or two every couple of days, you will probably be fine. You’re right that prior to implantation, the embryo isn’t in contact with your bloodstream. Even afterward, drinking small amounts and slowly allows you to metabolize most of the alcohol before it can affect the embryo. Once you know you’re pregnant (i.e. your embryo is implanted and its systems are in the critical early stages of development) avoiding alcohol is a wise course during the first trimester when there may be an increased risk of miscarriage associated with light drinking (the evidence is mixed – she analyzes it on pp. 48-49 of her book).
During the second and third trimesters, drinking small amounts slowly has not been associated with harm to the fetus in studies with well-designed methodologies. Many people and organizations will insist vehemently that “No amount of alcohol has been shown to be safe in pregnancy” and that even light drinking in pregnant women is associated with behavioral and cognitive problems in children born to them. There was a recent thread where a whole bunch of bees were up in arms about it, insisting “9 months isn’t that long, I can give up anything for 9 months for the good of my baby,” etc., subtly implying that anyone who didn’t was selfish or a bad mother. However, there’s a difference between sacrificing something because the evidence supports it, and sacrificing something because of social pressure/social convention/mommyjudging when the evidence doesn’t back it up. A lot of the logic behind categorical prohibitions of alcohol in pregnancy is rooted in the fear that if you tell women “a little is okay,” they will blow it out of proportion and decide that “a little” is a lot more than it really is. That’s pretty patronizing and infantilizing toward women – we don’t suddenly lose the ability (or the *right*) to make good decisions for ourselves and our babies just because we’re pregnant. True, some women don’t make good decisions when they’re pregnant (and they are the ones the categorical prohibition is trying to get to) but, seriously, those are the women who are probably going to make bad decisions no matter what. Really restrictive approaches to foods, drinking, etc., in pregnancy frequently end up putting unnecessary limits on careful, well-intentioned women (and causing them unnecessary stress) way more than they improve the behavior of women inclined to make bad choices.
The main evidence for harm associated with light drinking comes from studies with methodological flaws. One of the biggest and most often-cited is a 2001 Pediatrics study that had a pretty major methodological issue: of the women who participated in the study, 18% of the non-drinkers and 45% of the women who had one drink per day reported cocaine use during their pregnancies! So, yeah, if nearly *half* of your “light-drinking” group also happens to be using cocaine, it’s hardly surprising that they would see cognitive and behavioral problems in a statistically significant proportion of those children, and the study design makes it impossible to separate what the alcohol contributed as compared with cocaine – but other studies that have shown no association between light drinking and behavioral problems means it’s reasonable to infer that anything alcohol contributes on its own is comparatively small. The study authors concluded (as they had to, based on the outcomes of their data) that even one drink per day was associated with behavioral problems, but they left out that one little key detail about the cocaine. Umm, yeah.
The key thing Oster points out is that it’s *always* smart to avoid binge drinking, even just a few times, during pregnancy. When a lot of alcohol hits your system really fast, that hits the baby’s system in amounts greater than its immature liver can handle. During early pregnancy this raises the risk of miscarriage and impacts developing systems. During the later trimesters, binge-drinking contributes to cognitive and behavioral problems.
So, don’t go doing shots (or cocaine!), but don’t worry about sipping a little wine.