This post, on the CDC recommendations about alcohol consumption by fertile women not using contraception was first published last February. It's worth re-posting, given that what it describes is a part of a longer-term trend.
1. The USAToday's Summary of New CDC Recommendations
The Big Brother has arrived! According to the USAToday:
Women of childbearing age should avoid alcohol unless they're using contraception, federal health officials said Tuesday, in a move to reduce the number of babies born with fetal alcohol syndrome.
“Alcohol can permanently harm a developing baby before a woman knows she is pregnant,” said Anne Schuchat, principal deputy director of the Centers for Disease Control and Prevention. “About half of all pregnancies in the United States are unplanned, and even if planned, most women won’t know they are pregnant for the first month or so, when they might still be drinking.
"The risk is real. Why take the chance?” Schuchat asked.
The CDC estimates 3.3 million women between ages 15 to 44 are at risk of exposing a developing fetus to alcohol because they drink, are sexually active and not using birth control. Even when women are actively trying to get pregnant, three in four continue drinking after they stop using birth control, according to the CDC report.
There is no known safe level of alcohol at any stage of pregnancy, according to the CDC. The American College of Obstetricians and Gynecologists recommends women abstain completely from alcohol while pregnant.
The bolds are mine. Read the first bolded sentence and then the second bolded sentence. Notice any difference? The hook in the article tells us that all women not using contraception who belong to a usually fertile age group should stop drinking, for the sake of future babies (whether planned or completely imaginary, and even if they will be born to someone else). Even Lesbians, hermits, nuns, other celibate individuals and infertile people should abstain from alcohol! Any woman might accidentally fall upon a penis, I guess.
Now imagine the Pre-Pregnancy Police coming for you if you try to get a drink and don't have enough wrinkles to prove your new legal drinking age! (1) Bartenders and other volunteers might refuse to serve you that glass of wine or at least first demand to know if you are on the pill, and then decide if you are allowed to drink.
The Pre-Pregnancy Police doesn't yet exist. But the Pregnancy Police, in the form of not only actual police but also concerned volunteers is a real thing and a real pest for pregnant women. I guess one advantage of this new recommendation is that now those helpful strangers can pester all younger women equally and not just the ones who are visible pregnant.
After writing that rant about the USAToday summary I read what the CDC actually says:
An estimated 3.3 million women between the ages of 15 and 44 years are at risk of exposing their developing baby to alcohol because they are drinking, sexually active, and not using birth control to prevent pregnancy, according to the latest CDC Vital Signs report released today. The report also found that 3 in 4 women who want to get pregnant as soon as possible do not stop drinking alcohol when they stop using birth control.
Alcohol use during pregnancy, even within the first few weeks and before a woman knows she is pregnant, can cause lasting physical, behavioral, and intellectual disabilities that can last for a child’s lifetime. These disabilities are known as fetal alcohol spectrum disorders (FASDs). There is no known safe amount of alcohol – even beer or wine – that is safe for a woman to drink at any stage of pregnancy.
Bolds are mine.
That is not the same as the first sentence in the USAToday story. I wish newspapers didn't promote shitty journalism.
2. The CDC Recommendations. On Statistics And Medical Studies.
But even more I wish that the people at CDC had a better understanding of statistics, more transparency about what medical research actually shows and doesn't show. I also wish that they had hired someone who would have edited the writing in this sentence:
An estimated 3.3 million women between the ages of 15 and 44 years are at risk of exposing their developing baby to alcohol because they are drinking, sexually active, and not using birth control to prevent pregnancy.
Those 3.3 million women don't all have "a developing baby". They are potentially at risk for becoming pregnant. Those two are very different things, and what is developing during any resulting pregnancy is not called a baby until it is born.
For the statistical problems, consider this quote that was used in the USAToday article as well as in the original CDC report:
About half of all pregnancies in the United States are unplanned, and even if planned, most women won’t know they are pregnant for the first month or so, when they might still be drinking.
That half of all pregnancies in the United States are unplanned does NOT mean that every woman has a 50% chance of having an unintended pregnancy! Yet all public health announcements aimed at fertile women seem to assume that the 50% frequency difference applies to every single fertile woman, even those who don't have heterosexual intercourse.
The actual situation is quite different, as this Guttmacher Institute graph shows:
I quote from the graph: The two thirds of US women at risk of unintended pregnancy who practice contraception consistently and correctly account for only 5% of unintended pregnancies.
I suspect that the CDC researchers who wrote the recommendation did take that Guttmacher information into account, because the recommendation doesn't extend to women who use reliable contraception. But the USAToday made a hash of it all and the CDC still parrots the statement without giving that sentence I bolded.
Even the more moderate statement from the CDC is not moderate when it comes to certain hidden assumptions about what various groups of women can be asked to sacrifice and for what types of reasons. To see why that is the case, let's talk about the medical evidence on fetal alcohol syndrome (FAS).
This quote from the CDC summary is a good starting point:
There is no known safe amount of alcohol – even beer or wine – that is safe for a woman to drink at any stage of pregnancy.
The quote is about alcohol consumption during pregnancy. But because we now regard all childbearing-age women as potentially pregnant or pre-pregnant, what with all those unplanned pregnancies and the fact that a woman might not know right away that she is pregnant, statements about pregnancy have a tendency to get generalized to more and more fertile-age women.
But that's not the important point I want to make about the quote. Rather, it is this:
It is certainly true that studies have not established what level of alcohol consumption might be safe in pregnancy. But that's because studies haven't addressed that question, as far as I know!(2) Most studies are about severe alcohol abuse during pregnancy, the kind we'd usually associate with alcoholism.
That studies have not established a safe level of alcohol consumption during pregnancy doesn't have to mean that no such level exists. Otherwise it would be hard to explain why almost every single citizen in the earlier Spanish, French and Italian generations didn't suffer from FAS. Wine in those countries is consumed routinely, and even pregnant women in the past drank it, most likely daily.
So where did that no-safe-amount recommendation come from if it didn't come from studies which literally found that even one drop of alcohol is dangerous during pregnancy?
The initial recommendation seems to have been born from the assumption that telling pregnant women not to drink any alcohol is a recommendation with small psychological costs to the woman and potentially large benefits to the children they are going to give birth to. What's the sacrifice in not having any wine or beer or cocktails for nine months? No sacrifice at all, if it guarantees that the baby will be born without FAS.
But now that recommendation is extended to all sexually active women who don't use contraception. The costs of that new recommendation are much higher, because the length of recommended abstention from alcohol use is much longer and because many more women are affected by it. Those costs include the loss of any health benefits which might come from, say, drinking red wine. They also include the additional psychological costs from harassment by the Pre-Pregnancy Police volunteers.
I suspect that pediatricians, as a group, are not going to think of those costs to the affected women. They understandably focus on the children with FAS and aren't too bothered about the rights of fertile-age women in general or about the racial, ethnic and income differences which correlate with unintended pregnancies (3)
To compare those concerns (theirs: children born with FAS, and mine: as many equal rights for women and men as is biologically possible (4)), we need to know how common FAS is. If even small amounts of alcohol during pregnancy can cause it, the condition should be very common.
Indeed the USAToday summary gives frightening prevalence figures:
Alcohol use during pregnancy "can cause lasting physical, behavioral and intellectual disabilities that can last for a child’s lifetime," according to the CDC. These disabilities, known as fetal alcohol spectrum disorders, affect up to one in 20 school children, according to the CDC.
Bolds are mine.
But that is not the same as FAS. The CDC gives the following figures for that:
Using medical and other records, CDC studies have identified 0.2 to 1.5 infants with FAS for every 1,000 live births in certain areas of the United States.1 The most recent CDC study analyzed medical and other records and found FAS in 0.3 out of 1,000 children from 7 to 9 years of age. 2
Studies using in-person assessment of school-aged children in several U.S. communities report higher estimates of FAS: 6 to 9 out of 1,000 children. 3,4Those figures suggest that FAS is a relatively rare condition, something that could very well be associated with severe alcohol consumption during pregnancy.
Fetal alcohol spectrum disorders (FASD) is a term which covers a mix of physical, behavioral and learning problems in a child when those are believed to have been associated with maternal drinking during pregnancy. The fetal alcohol syndrome FAS is the most serious form of FASDs. The fourth reference in the above quote tells us this about the prevalence of FASD:
Regarding all levels of FASD, we estimate that the current prevalence of FASD in populations of younger school children may be as high as 2-5% in the US and some Western European countries.Compare that to the USAToday quote which only gives us the 5% figure and omits the "may be" part. That is nitpicking, my friends. But the point is that the USAToday article chose to go for the maximum impact, and the highest possible numbers, even switching from FAS to FASD in order to get there.
So what might cause FASD? The third study referenced in the CDC quote above tells us the variables which were the best predictors for finding FASD in a study of young school children in the Midwest:
The most predictive maternal risk variables in this community are late recognition of pregnancy, quantity of alcoholic drinks consumed 3 months before pregnancy, and quantity of drinking reported for the index child's father.Hmm. The authors of the study write quite a bit about the difficulty of getting honest answers from mothers of children with FASD, for fairly obvious reasons. If those mothers under-report alcohol consumption during pregnancy, then both the quantity of alcoholic drinks consumed prior to pregnancy and the woman's partner's consumption of alcohol could be proxies (indirect measures) for that under-reported drinking during pregnancy. (5)
Or the mother's drinking during the three months preceding pregnancy and the alcohol consumption by the child's father could both be real causes for the birth of a child with FASD. But we are not telling fertile men who have sex without contraception to abstain from all alcohol use.
And that's the main reason for all this nitpicking: The way we think of reproduction is still almost 100% female. As one article notes:
In evaluating why there is still so little research about how men and their sperm matter for reproductive outcomes (Daniels 2006), we point to the widespread and stubborn belief among our respondents in the overriding significance of women’s role in reproduction. As one nurse involved with the CDC’s PCHHC Initiative put it, “We can’t get around it; it’s not the man having the baby.”
That this is true even among those who focus on the period before conception, which ostensibly offers an opening for thinking in more egalitarian ways, suggests the intransigence of gendered ways of figuring reproductive equations. The end result is deeply gendered knowledge about reproduction that simultaneously leaves open social and clinical questions about men while reinforcing women’s part of the equation through research and medical practice.Bolds are mine.
That last sentence makes an important point:
As long as research into fertility and conception assumes that only the female partner in a heterosexual couple matters, research into fertility and conception will focus on only the female partners. Thus, we end up accumulating more and more research results on women. Those, in turn, will reinforce the initial assumption that the men who become fathers just hand over some perfect sperm and women then sometimes make a mess out of the ultimate child by behaving badly.
Perhaps more importantly, such research biases can slow down the process of learning about all the possible causes of various congenital conditions.
Try this thought experiment: Suppose that you have the data the CDC had and decide that it's sufficient to make a recommendation about people's pre-conception behavior. You have several choices.
1. You could recommend that all fertile women who are sexually active in a heterosexual relationship and who consume alcohol should use reliable contraception, and that all couples who plan to become pregnant should abstain from alcohol consumption.
2. You could recommend that all fertile men should use reliable contraception if they have vaginal sex with women who don't use contraception and who consume alcohol.
3. Or you could recommend that all childbearing age women who have unprotected heterosexual intercourse should abstain from alcohol for as many years as those conditions remain in place.
There are probably further choices than those I have listed. The point is that the one the CDC picked tells us something about the underlying beliefs and values, about who is assigned the responsibility and for what.
None of those recommendations is terribly likely to work in practice. But I think recommendation number one has the highest odds of working at all. So why didn't the CDC go with that?
And why did the USAToday choose to publish an exaggerated version of the CDC recommendation?
And why did the USAToday choose to publish an exaggerated version of the CDC recommendation?
1. Earlier guidelines have gone in the same direction, which is to regard all women as potentially pregnant until menopause. An example can be found here.
2. This article discusses several studies on the question of moderate alcohol consumption during pregnancy. Keep in mind that "pre-pregnancy" is not the same thing as pregnancy. A few quotes:
A study cited by the pediatricians' group found the risk of having a baby with growth retardation goes up even when a woman has just one alcoholic drink a day.and
Over the years, some studies have found that a moderate amount of drinking during pregnancy is not linked to cognitive or behavioral issues. A 2010 study found that children of mothers who drank one or two glasses of alcohol a week while pregnant had no problems with behavioral or intellectual development by the time they turned 5.
A study three years later found that the children of mothers who drank three to seven glasses of alcohol a week did not have trouble with balance tasks at age 10, a sign of neurological development.
For proper analysis of what it means to say the "women drank alcohol" during, say, first trimester of their pregnancies I'd need to read all the original papers. "Drinking alcohol" is a qualitative statement, not a quantitative statement, and doesn't tell us how much and how often the women in the studies consumed alcohol. Note that having "just" one alcoholic drink a day while pregnant isn't what I'd call light drinking, yet the study the quote discusses seems to interpret it that way.But doctors warn that even if these studies are accurate, every baby and mother will react differently to alcohol, and so the safest thing is to not drink at all.
According to Wikipedia some experts define moderate wine consumption for women as less than one 5-US-fluid-ounce (150 ml) glass of wine per day. Moderate wine consumption for men is defined as less than two 5-US-fluid ounce glasses of wine per day.
The last sentence in the above quote is the usual way the ban on alcohol is justified: Because we don't know, better be safe rather than sorry.
As I have noted in the body of the post that may be good advice during pregnancy and while actively trying to get pregnant. The gender-political question is how far into the lives of all women those recommendations should reach and what strength we require from the research results before wholesale behavior modification is demanded.
3. From the Guttmacher Institute Report:
Unintended pregnancy rates are highest among poor and low-income women, women aged 18–24, cohabiting women and minority women. 
4. This is an artificial juxtaposition, of course, because I obviously care about both issues. I want FAS to be prevented, but before we ask three million women to become permanent teetotalers it's worth asking whether that, indeed, is the best, cheapest and least restrictive policy we could find. Why not tell them to use contraception if they want to drink?
Thus, questions about what the research actually tells us matter. That in most decisions of this kind (such as setting speed limits on highways) we look at alternative and potentially opposing goals should matter. And the gender politics of medical research should also matter: If research focuses on maternal drinking and largely ignores paternal pre-conception drinking, then the recommendations will only apply to women and not to men.
5. The mother's alcohol consumption in the three months before pregnancy could be a more honest reflection of her average alcohol consumption, even while pregnant, because the stigma attached to drinking during pregnancy makes under-reporting likely.
The father's alcohol consumption could correlate with the mother's alcohol consumption if partners tend to have similar drinking habits. In one sense, both of these may have been found as predictive because they are correlated with the quantity of alcohol the mother may have actually consumed while pregnant. All this is speculation, of course.