Caffeine Safety in Pregnancy
My first trimester of pregnancy coincided exactly with the last three months before my book deadline. I was lucky to have only mild nausea during this time, but I was really, really tired, especially in the afternoon. I tried hard to get enough sleep at night, but my body also seemed to want a 2-hour nap after lunch, when I just couldn’t stay awake, much less think and write. Pre-pregnancy, I responded to a post-lunch slump by pouring myself a cup of coffee or, even better, spending the afternoon at my favorite coffee shop, where a latte and the people around me helped keep me focused for a productive afternoon. A cup of herbal tea in the same atmosphere just made me want to curl up in one of the comfy chairs and take a nap, even as my caffeine-fueled coffee shop friends typed energetically around me.
But now I was pregnant, after 18 month of trying and several miscarriages, and I wanted to do all I could to minimize the risks of losing this pregnancy. In previous pregnancies I’d just given up most caffeine, and that wasn’t that hard to do. In this one, I was more afraid than ever of a miscarriage, but I also needed the caffeine boost more than ever to finish my book. I wanted to know what the research says about the safety of caffeine in pregnancy so that I could make an informed decision about whether to consume caffeine, and if so, how much.
In her book, Expecting Better, Emily Oster includes an excellent discussion of caffeine in pregnancy. I consulted this for a quick answer to my question, and her analysis of the research on this topic helped me feel comfortable strategically drinking a little coffee in the afternoon. However, as much as I respect and highly recommend Oster’s book, I’ve also found that my approach to risk in pregnancy is a bit more conservative, and as soon as I had the chance, I wanted to look at the studies myself.
Major sources of caffeine
Caffeine is found in coffee, soft drinks, tea, energy drinks, chocolate, and some medications and weight loss supplements. The table below gives caffeine content of many of these products so that you can estimate your own caffeine intake. Note that caffeine content of coffee and tea can vary quite a bit depending on type and preparation. For example, an 8 oz. serving of drip coffee has been measured to have anywhere between 70 and 280 mg of caffeine. My coffee shop makes every espresso drink with two shots of espresso, and they’re proud of their strong drinks. (My standard pregnancy order there is a half-caf latte.)
*This is by no means a complete list of medications and weight loss supplements that contain caffeine. Consult a health care provider about the safety of all medications and supplements in pregnancy. Source: Anderson et al. Caffeine’s implications for women’s health and survey of obstetrician-gynecologists’ caffeine knowledge and assessment practices. Journal of Women’s Health 18, 1457–1466 (2009).
Why be concerned about caffeine use in pregnancy?
We know that caffeine crosses the placenta from the maternal to fetal bloodstreams, and the fetus can only slowly metabolize, or break down, caffeine. On top of this, maternal metabolism of caffeine is slowed in pregnancy, so caffeine lingers in mom’s bloodstream for longer time. The half-life of caffeine, or the time it takes for half of the caffeine to leave the bloodstream, is estimated to be 2-6 hours in older children and adults, 10-20 hours in pregnant women, and 4 days in newborn babies. (From a practical standpoint, this means that a cup of coffee might keep you awake longer than it did prior to pregnancy.)
So we know that when we consume caffeine during pregnancy, it enters the fetal circulation, but is this necessarily harmful? Animal studies show that high intakes of caffeine can cause birth defects, but these are very high intakes, and there’s no evidence of this occurring in humans. The main concern with caffeine in pregnancy is that it might increase the risk of miscarriage. However, this is one of those questions where there are lots of studies but also lots of confusion, making it an interesting case study in the limitations of research, particularly when it comes to pregnancy.
Can caffeine cause miscarriage?
There have been many studies of this question, but they all have serious limitations. A 2013 Cochrane review of caffeine safety in pregnancy only found one study that met its criteria, and it didn’t even address the miscarriage question. A 2004 review of papers published since 1966 identified 15 relevant studies but concluded that their methodological flaws were so great that they essentially couldn’t draw any conclusions from them. Why is it so difficult to study and answer this question?
The biggest problem is one that plagues almost all of pregnancy and parenting research: it is nearly impossible to do randomized controlled trials. Instead, we’re stuck with observational data. A cohort study, for example, would measure women’s caffeine intake early in pregnancy (usually by asking the women to estimate how much they consumed) and then track which women ended up miscarrying. If women who miscarried had consumed more caffeine than those who didn’t, then we could conclude that there is a correlation or an association between caffeine intake and miscarriage. However, this is not evidence that caffeine actually causes miscarriage. There are a number of problems with drawing conclusions from observational studies on caffeine use in pregnancy:
- Confounding factors. In a randomized controlled trial, we could compare two or more groups of women who were otherwise very similar except in how much caffeine they consumed. But in the real world, and in observational studies, a variable like caffeine consumption doesn’t exist in isolation. For example, studies have found that women who consume more caffeine in pregnancy are also more likely to be smokers, to drink alcohol in pregnancy, and to be older – and any of these factors could also increase miscarriage risk. Researchers usually try to statistically control for these confounding factors, but we can never be sure if the statistical methods are adequate or if the information is accurate. (For example, some women might not tell the truth about smoking in pregnancy.)
- Pregnancy symptoms. The biggest confounding factor is probably that pregnancy itself can affect caffeine consumption. Many women feel nauseous in the first trimester, and coffee, with its strong taste and odor, can lose its appeal during this time. Furthermore, women with greater symptoms of nausea and vomiting – the least likely to want a strong cup of coffee – are also less likely to miscarry. These symptoms seem to be a signal that the pregnancy is going well. Women who have a “missed” miscarriage, in which the fetus has stopped developing but mom’s body doesn’t naturally expel it right away, may find that their symptoms of nausea and vomiting have subsided before the miscarriage is discovered. Thus, women who consume more caffeine in pregnancy may already be at higher risk for miscarriage or may have already had one. While some studies have attempted to correct for this problem by asking women about their pregnancy symptoms, one woman’s “severe nausea” may be another’s “mild nausea.” It’s hard to quantify such a subjective variable or to account for changes in pregnancy symptoms over time.
- Recall bias. The problem here is that most studies have asked women after a miscarriage may have occurred how much caffeine they consumed in early pregnancy, and women who have already experienced the trauma of miscarriage may remember their pregnancies differently from those who didn’t. I remember just as many details about a miscarried pregnancy 18 months ago than I do about the first trimester of this pregnancy, which ended just a few weeks ago. I’ve gone over the details of that lost pregnancy in my mind so many times that those memories are vivid and strong. To reduce recall bias, researchers need to ask women about their caffeine consumption as soon as they know they’re pregnant, before a miscarriage might occur, but most studies have not been conducted this way.
All of these sources of bias push the data in one way, tending to overestimate the association between miscarriage and caffeine consumption.
Two prospective cohort studies published in 2008 claimed to have overcome many of these limitations, but although their methods were improved, there was still lots of room for bias. Even more confusing, they came to different conclusions. One study led by David Savitz of Mount Sinai School of Medicine found no relationship at all between caffeine consumption before and during early pregnancy in more than 2400 women. The second study, led by Xiaoping Weng of Kaiser, tracked more than 1000 pregnant women and concluded that there was an increased risk of miscarriage associated with consuming more than 200 mg of caffeine per day, but not less than this. Interestingly, women with a history of miscarriage showed no association between caffeine and miscarriage, providing some reassurance for women like me and suggesting that in these cases, there are more important factors contributing to miscarriage risk.
Finally, a 2010 study led by Anna Pollack of Johns Hopkins did a nice job of avoiding many of the limitations mentioned earlier by asking women to keep daily diaries of caffeine, alcohol, and tobacco use during a full year, beginning before pregnancy. Like the Savitz study, this study found no relationship between caffeine and becoming pregnant or miscarriage. The problem is that it only included 79 women, 11 of whom miscarried, a sample size too small to answer this question with much certainty. None of these 3 studies included many women consuming more than 300-400 mg/day, limiting their ability to look at risks of those higher caffeine intakes.
In a 2010 Committee Opinion, ACOG concluded that 200 mg of caffeine per day or less probably doesn’t increase the risk of miscarriage, mostly based on the studies by Savitz and Weng. Given all of the potential sources of bias and the studies finding no association at all, this is a conservative recommendation. Chances are, consuming more caffeine (to an extent) probably won’t increase your chances of miscarriage, but ACOG errs on the side of caution.
Can caffeine cause preterm birth or affect birth weight?
For this question, we actually have good data. A clever randomized controlled trial recruited pregnant women who were already drinking at least 3 cups of coffee per day. The researchers randomized the women to two groups, and starting around 18 weeks of pregnancy, they gave one group regular instant coffee and the other decaf instant coffee (without telling them which they were receiving). The women were asked to drink this coffee at home but were allowed to have other sources of caffeine as they wished. The decaf group ended up consuming 177 mg/day and the caffeinated group consumed 317 mg/day. Despite this difference in caffeine intake, there was no difference in preterm births, birth weights, or Apgar scores. Thus, at least within this range of caffeine intake, there doesn’t appear to be a risk of these outcomes.
Information means better choices and less guilt
Some women will look at this research and decide they’ll just avoid caffeine altogether, as I did in previous pregnancies. If, in this pregnancy, the research showed a real risk to consuming a little caffeine, I’d certainly continue to avoid it. But the research shows that consuming 200 mg per day or less is not a risk, and more is probably okay. In my case, a cup of coffee or so per day was helpful in allowing me to remain productive throughout my first trimester. Regardless of our individual choices, it is important to be honest about the risks (or lack thereof) of caffeine in pregnancy, particularly for women who do experience pregnancy loss. Many wonder if it was their fault, and they deserve to know that their miscarriage most likely had nothing to do with caffeine.
Have you consumed caffeine in pregnancy? What information did you use to make this decision?