This is the first of a series on how Oster is wrong on drinking in pregnancy in her 2013 book Expecting Better: Why the Conventional Pregnancy Wisdom is Wrong and What You Really Need to Know.
For this part, we’ll focus on how and why her biological model of ethanol metabolism is incorrect. This premise forms part of her argument on why it’s acceptable to drink 1-2 drinks* a week in the first trimester and more later in pregnancy.
Oster writes,
To understand why there is a difference between excessive drinking and moderate or light drinking, it’s useful to think a little bit about how the biology works. Many women seem to think that when they drink, that glass of wine is channeled directly to the fetus. People correctly note that you would not give your infant a glass of wine, so why would you give your fetus one? Needless to say, this is not really how it works. When you drink, alcohol enters your digestive system and is passed into your bloodstream. Your liver processes the alcohol into a chemical called acetaldehyde and then into acetate. The acetaldehyde is toxic to other cells, and depending on how quickly you drink, it can remain in your bloodstream. You share your blood with your baby through the placenta; acetaldehyde, which remains in your bloodstream, is therefore shared with the fetus. Your baby actually can process some alcohol, but not as much as an adult (obviously). If too much acetaldehyde is passed to the baby, it can get into his tissues and impact development. When you drink slowly, you metabolize much of the alcohol before it would get to the fetus. If you drink quickly, your liver cannot keep up and toxins are passed to the fetus. This is why binge drinking is so bad, but it also illustrates why negative effects of light drinking do not follow directly from negative effects of heavy drinking.
The bolded statement is, quite simply, not in any way true when it comes to alcoholic drinks, and particularly not in the first trimester. You do not meaningfully metabolise 1 drink before it makes it to the fetus. We know this because your blood alcohol level rises; blood alcohol is a measure of how much ethanol is actually in the blood. Its very presence in the blood means it hasn’t been metabolised yet and is free to diffuse into fetal tissue.
Oster is completely correct that a woman drinking 1 glass of wine isn’t like a fetus drinking 1 glass of wine itself. For one, body size! A first trimester embryo/fetus has less mass than the mass of the ethanol in the wine itself so this is trivially true.
We also have enzymes in our intestinal tract (ADH4) that digest some of the ethanol before it reaches the blood stream. However, these only digest a very small fraction of the ethanol we consume. The vast majority(92-98%) is metabolised by blood passing through the liver.
In a more realistic scenario, after 1 drink, it takes about a half hour for maternal BAC to peak—during that 30 minutes, that ethanol is continuously diffusing into both maternal and fetal tissue. In the first trimester, the placenta is yet to form and so no placental filtering is occurring.
The rise in BAC indicates it’s being absorbed faster than it’s being metabolised; if it was being metabolised faster, it’d be dropping. Some of that ethanol is being metabolised, but that’s already calculated in with maternal BAC—if it hadn’t been, BAC would be even higher. So it’s not enough to counteract the rise in BAC.
But let’s continue on with her explanation; Oster claims maternal acetaldehyde, a metabolite of ethanol, gets “passed to the baby” (this is almost certainly false but we’ll entertain it). How much alcohol, exactly, can you metabolise into acetaldehyde, and how quickly? If acetaldehyde getting passed to the baby is the problem, we need to calculate at what threshold it will build up.
Let’s calculate!
Steel-manning the maternal acetaldehyde hypothesis
One standard drink (in the US, that is) is defined as containing 14 grams of alcohol. This is actually a lot of alcohol! The liver does this pretty slowly. For the average metaboliser, hepatic enzyme is saturated at 15–20 mg/100 mL, which translates to a reduction of .015%-.02% BAC per hour). (From here on out I’ll use the higher .02% value; for some people, like alcoholics, it can be as high as .03%)
To steel-man Oster’s mechanism for a minute, in order for acetaldehyde to build up, the rate at which acetaldehyde is produced by ADH must exceed the rate at which acetaldehyde is converted into acetic acid. At what rate of acetaldehyde production must that occur?
As I said before, that’s actually very low. Lower than binge drinking, which legally is .08% BAC. It turns out that at .02% BAC, the ADH enzyme is full saturated (meaning it’s working at maximum rate.) So above ~.02% BAC, acetaldehyde production is constant.
But it turns out that ALDH, the enzyme that converts acetaldehyde to acetic acid, has a much higher capacity than ADH does in most women. This is by design; because acetaldehyde is so dangerous, ADH is prevented from producing more acetaldehyde than ALDH can handle. In some women, ALDH may be defective, meaning acetaldehyde could build up—but in this case essentially any amount of drinking would be dangerous for these individuals.
If acetaldehyde can build up in an individual, it must at the very least be happening any time she goes above .02% BAC because it would be produced at a constant rate. Using the Widmark formula again, let’s see what proportion of women will have blood alcohol levels above that with 1 drink.
This is a simple toy model that creates a normal weight distribution from the CDC values for average weight and STD dev of American women, a value of .02% BAC clearance rate, and the Widmark formula.
As you can see from the graph, the vast majority of women will be ADH saturated (meaning acetaldehyde production is at its maximum) after one drink (orange > .02% BAC; pink > .08% BAC). Only the women in blue are below the threshold. However, you may point out this is a static graph that shows what happens after 1 drink; if acetaldehyde production were to fall quickly enough, it’s perhaps not relevant to the embryo/fetus.
In this next graph I calculate how many hours at minimum are spent at a BAC > .02% given the number of drinks and weight. Widmark formula does not work at BAC <.02% as actually clearance occurs more slowly below this, so the numbers are categorically an underestimate for the total duration of ethanol exposure.
You can see that a half a drink is under the saturation threshold for most women except for those under 115 lbs. This also demonstrates how sensitive BAC is according to weight. For a 115 lb woman, just 2 drinks is binge drinking. Even if it is the case that only a BAC over .08% is problematic for pregnancy, 1-2 drinks a week is pretty close to that threshold for lower-weight woman.
For the average weight woman who is 170 lbs, she’ll reach the legal limit at 3 drinks and spend 4 hours before her BAC drops below .02%. At 1 drink, that will only take a half hour. However, a 125 lb woman having that same 1 drink will take more than an hour to get below the threshold. For two drinks, it will take 3.5. For 3 drinks, 6 full hours.
If it is the mechanism then it would only apply in rare cases where a person’s ethanol metabolism is dysfunctional. And for those people, any amount of drinking would be dangerous—the safe threshold at most weights would likely be well less than half a drink. (There are, in fact, a large number of people with defective ALDH genes but behaviourally they don’t drink as much because of the unpleasant side effects—you might know it as “alcohol flush reaction.”) For the rest of us, though, we could in theory drink as much as we wanted, as long as we knew our ALDH was fine. (Interestingly, defective acetaldehyde metabolism might itself be an adaptation against ethanol consumption, because it’s so unpleasant people with it drink very little if at all.)
But, it seems likely it’s not exposure to maternally produced acetaldehyde that’s the problem at all.
But basically, this tells us nothing except more is bad which we already knew. There’s no threshold effects in this model at all! Oster’s idea that a mechanism is both a) known and also b) gives us some sort of cut-off, is wrong.
Why might direct ethanol exposure in the embryo or fetus be bad? We don’t actually know for sure. But there are some compelling new experiments that suggest in the first trimester at least, it’s actually fetally produced acetaldehyde, not the mother’s, when exposed directly to ethanol. It’s thought that acetaldehyde competes with retinaldehyde for the enzyme RALDH2 which is critical for embryonic development in the first trimester. But we don’t know precisely what level of ethanol exposure to fetal tissue causes this—the amounts used in the study weren’t biologically realistic (.5% EtOH, which is an order of magnitude higher than typical BACs. The LD50 of ethanol is .4% BAC). The use of biologically realistic ethanol concentrations would make for an interesting follow-up.
However this is still an area where the research is unsettled; there could be other direct effects on fetal tissues of ethanol. The exact mechanism is unknown.
The Bottom Line
1) Oster is not correct that drinking slowly allows you to metabolise alcohol quickly enough to avoid exposing fetal tissue. In the first trimester diffusion of ethanol into fetal tissue occurs roughly concurrently with rise of and fall of maternal BAC.
2) If Oster was correct and maternal acetaldehyde build-up was the correct mechanism, her advice of 1-2 drinks a week is neither necessary nor sufficient to prevent FASD. In a hypothetical scenario where acetaldehyde metabolism was defective, even light drinking would cause maternal acetaldehyde to build up, not just binge drinking. In most women, acetaldehyde does not build up, even in a binge drinking scenario.
3) It’s likely not maternal acetyladhyde, but exposure to ethanol itself that causes FASD. The fetus is exposed to ethanol even in light drinking, because the amounts that are drunk recreationally are recent in our evolutionary history, and ethanol metabolism cannot work quickly enough to avoid exposing the fetus.
Footnotes
*Note that the definition of 1 standard drink differs substantially from country to country. I use US numbers, data, and definitions throughout for consistency as Oster is writing for an American audience.
Contra Oster on Alcohol in Pregnancy. Part 1. The pharmacokinetics of alcohol metabolism
This is the first of a series on how Oster is wrong on drinking in pregnancy in her 2013 book Expecting Better: Why the Conventional Pregnancy Wisdom is Wrong and What You Really Need to Know.
For this part, we’ll focus on how and why her biological model of ethanol metabolism is incorrect. This premise forms part of her argument on why it’s acceptable to drink 1-2 drinks* a week in the first trimester and more later in pregnancy.
Oster writes,
The bolded statement is, quite simply, not in any way true when it comes to alcoholic drinks, and particularly not in the first trimester. You do not meaningfully metabolise 1 drink before it makes it to the fetus. We know this because your blood alcohol level rises; blood alcohol is a measure of how much ethanol is actually in the blood. Its very presence in the blood means it hasn’t been metabolised yet and is free to diffuse into fetal tissue.
Oster is completely correct that a woman drinking 1 glass of wine isn’t like a fetus drinking 1 glass of wine itself. For one, body size! A first trimester embryo/fetus has less mass than the mass of the ethanol in the wine itself so this is trivially true.
We also have enzymes in our intestinal tract (ADH4) that digest some of the ethanol before it reaches the blood stream. However, these only digest a very small fraction of the ethanol we consume. The vast majority (92-98%) is metabolised by blood passing through the liver.
To drink so slowly that no alcohol makes it into the maternal blood stream? I cannot provide an exact estimate on how little ethanol or how slowly that would be because no one has bothered to study such tiny amounts of consumption. I found one study where they had volunteers drink one beer very slowly—over 2 hours vs 1 hour or half hour. Peak BAC was only reduced for the 2 hour group but the fall in BAC was also slower because consumption of 20ml/5min still grossly outpaced ethanol metabolism. Any biologically plausible ethanol consumption amount that could show no rise in BAC would be necessarily have to be much, much less than one beer over two hours and probably in the “trace amounts” category. We can discard this idea entirely unless Oster imagines women are drinking their one glass of wine in 100 μL aliquots over a time scale of days.
In a more realistic scenario, after 1 drink, it takes about a half hour for maternal BAC to peak—during that 30 minutes, that ethanol is continuously diffusing into both maternal and fetal tissue. In the first trimester, the placenta is yet to form and so no placental filtering is occurring.
The rise in BAC indicates it’s being absorbed faster than it’s being metabolised; if it was being metabolised faster, it’d be dropping. Some of that ethanol is being metabolised, but that’s already calculated in with maternal BAC—if it hadn’t been, BAC would be even higher. So it’s not enough to counteract the rise in BAC.
But let’s continue on with her explanation; Oster claims maternal acetaldehyde, a metabolite of ethanol, gets “passed to the baby” (this is almost certainly false but we’ll entertain it). How much alcohol, exactly, can you metabolise into acetaldehyde, and how quickly? If acetaldehyde getting passed to the baby is the problem, we need to calculate at what threshold it will build up.
Let’s calculate!
Steel-manning the maternal acetaldehyde hypothesis
One standard drink (in the US, that is) is defined as containing 14 grams of alcohol. This is actually a lot of alcohol! The liver does this pretty slowly. For the average metaboliser, hepatic enzyme is saturated at 15–20 mg/100 mL, which translates to a reduction of .015%-.02% BAC per hour). (From here on out I’ll use the higher .02% value; for some people, like alcoholics, it can be as high as .03%)
To steel-man Oster’s mechanism for a minute, in order for acetaldehyde to build up, the rate at which acetaldehyde is produced by ADH must exceed the rate at which acetaldehyde is converted into acetic acid. At what rate of acetaldehyde production must that occur?
As I said before, that’s actually very low. Lower than binge drinking, which legally is .08% BAC. It turns out that at .02% BAC, the ADH enzyme is full saturated (meaning it’s working at maximum rate.) So above ~.02% BAC, acetaldehyde production is constant.
But it turns out that ALDH, the enzyme that converts acetaldehyde to acetic acid, has a much higher capacity than ADH does in most women. This is by design; because acetaldehyde is so dangerous, ADH is prevented from producing more acetaldehyde than ALDH can handle. In some women, ALDH may be defective, meaning acetaldehyde could build up—but in this case essentially any amount of drinking would be dangerous for these individuals.
If acetaldehyde can build up in an individual, it must at the very least be happening any time she goes above .02% BAC because it would be produced at a constant rate. Using the Widmark formula again, let’s see what proportion of women will have blood alcohol levels above that with 1 drink.
This is a simple toy model that creates a normal weight distribution from the CDC values for average weight and STD dev of American women, a value of .02% BAC clearance rate, and the Widmark formula.
As you can see from the graph, the vast majority of women will be ADH saturated (meaning acetaldehyde production is at its maximum) after one drink (orange > .02% BAC; pink > .08% BAC). Only the women in blue are below the threshold. However, you may point out this is a static graph that shows what happens after 1 drink; if acetaldehyde production were to fall quickly enough, it’s perhaps not relevant to the embryo/fetus.
In this next graph I calculate how many hours at minimum are spent at a BAC > .02% given the number of drinks and weight. Widmark formula does not work at BAC <.02% as actually clearance occurs more slowly below this, so the numbers are categorically an underestimate for the total duration of ethanol exposure.
You can see that a half a drink is under the saturation threshold for most women except for those under 115 lbs. This also demonstrates how sensitive BAC is according to weight. For a 115 lb woman, just 2 drinks is binge drinking. Even if it is the case that only a BAC over .08% is problematic for pregnancy, 1-2 drinks a week is pretty close to that threshold for lower-weight woman.
(Since most women don’t typically gain weight in the first trimester, I have used the weight distribution for non-pregnant women. There is also real data showing that severity of FASD is only associated with pre-pregnancy weight and not ameliorated by weight gain in pregnancy so I think this is a fine assumption.)
For the average weight woman who is 170 lbs, she’ll reach the legal limit at 3 drinks and spend 4 hours before her BAC drops below .02%. At 1 drink, that will only take a half hour. However, a 125 lb woman having that same 1 drink will take more than an hour to get below the threshold. For two drinks, it will take 3.5. For 3 drinks, 6 full hours.
Now after we’ve done all that steel-manning it might not matter. ALDH, the enzyme that converts acetaldehyde to acetic acid, actually has more capacity than ADH, which works more slowly. Acetaldehyde is really bad but even in binge drinking there aren’t typically detectable levels of it in the blood. This is why we actually have a lower metabolism of ethanol to acetaldehyde relative to acetaldehyde to acetic acid to begin with.
If it is the mechanism then it would only apply in rare cases where a person’s ethanol metabolism is dysfunctional. And for those people, any amount of drinking would be dangerous—the safe threshold at most weights would likely be well less than half a drink. (There are, in fact, a large number of people with defective ALDH genes but behaviourally they don’t drink as much because of the unpleasant side effects—you might know it as “alcohol flush reaction.”) For the rest of us, though, we could in theory drink as much as we wanted, as long as we knew our ALDH was fine. (Interestingly, defective acetaldehyde metabolism might itself be an adaptation against ethanol consumption, because it’s so unpleasant people with it drink very little if at all.)
But, it seems likely it’s not exposure to maternally produced acetaldehyde that’s the problem at all.
Rather, it’s probably direct exposure to ethanol itself—which is identical to the mother’s BAC, because in the first trimester fetal BAC roughly tracks maternal BAC. (In the second trimester and beyond fetal BAC does trail maternal BAC and maximum fetal BAC is reduced. However some downsides are that once this occurs it’s harder for ethanol to diffuse out—the fetal liver has to clear the ethanol itself, which means BAC remains elevated for longer than the mother as amniotic fluid is cycled, and the fetus’s ability to metabolise it is only about half that of the mother’s, i.e. .01% BAC/hr.)
But basically, this tells us nothing except more is bad which we already knew. There’s no threshold effects in this model at all! Oster’s idea that a mechanism is both a) known and also b) gives us some sort of cut-off, is wrong.
Why might direct ethanol exposure in the embryo or fetus be bad? We don’t actually know for sure. But there are some compelling new experiments that suggest in the first trimester at least, it’s actually fetally produced acetaldehyde, not the mother’s, when exposed directly to ethanol. It’s thought that acetaldehyde competes with retinaldehyde for the enzyme RALDH2 which is critical for embryonic development in the first trimester. But we don’t know precisely what level of ethanol exposure to fetal tissue causes this—the amounts used in the study weren’t biologically realistic (.5% EtOH, which is an order of magnitude higher than typical BACs. The LD50 of ethanol is .4% BAC). The use of biologically realistic ethanol concentrations would make for an interesting follow-up.
However this is still an area where the research is unsettled; there could be other direct effects on fetal tissues of ethanol. The exact mechanism is unknown.
The Bottom Line
1) Oster is not correct that drinking slowly allows you to metabolise alcohol quickly enough to avoid exposing fetal tissue. In the first trimester diffusion of ethanol into fetal tissue occurs roughly concurrently with rise of and fall of maternal BAC.
2) If Oster was correct and maternal acetaldehyde build-up was the correct mechanism, her advice of 1-2 drinks a week is neither necessary nor sufficient to prevent FASD. In a hypothetical scenario where acetaldehyde metabolism was defective, even light drinking would cause maternal acetaldehyde to build up, not just binge drinking. In most women, acetaldehyde does not build up, even in a binge drinking scenario.
3) It’s likely not maternal acetyladhyde, but exposure to ethanol itself that causes FASD. The fetus is exposed to ethanol even in light drinking, because the amounts that are drunk recreationally are recent in our evolutionary history, and ethanol metabolism cannot work quickly enough to avoid exposing the fetus.
Footnotes
*Note that the definition of 1 standard drink differs substantially from country to country. I use US numbers, data, and definitions throughout for consistency as Oster is writing for an American audience.