Depression during pregnancy is an important issue. Depression should not be ignored and depressed pregnant women deserve good treatment and care. Part of that good care, though, is providing them with full and correct information. I care for pregnant women taking antidepressants on a daily basis and too often they tell me that the only counseling they received about the medication was, “my doctor told me it’s safe in pregnancy.” In certain scenarios antidepressant use could be considered to be safer than the alternatives (for example, a patient who is suicidal when she stops her medications or one who would use cigarettes, alcohol, or illegal drugs). But these scenarios don’t mean that the chemicals (antidepressants) themselves are safe — only that they may be less risky than the alternatives.
Antidepressants simply cannot be considered to be “safe” in pregnancy and the five point argument below explains why this is the case:
- Serotonin is a neurotransmitter and cell-signaling molecule that is crucial for human development.
- The SSRI antidepressants are synthetic chemicals that disrupt the serotonin system.
- The SSRI antidepressants freely cross the placenta and enter the fetus.
- Animal studies show harm.
- Human studies show harm.
By antidepressants, I am referring here mostly to the selective serotonin reuptake inhibitors (SSRIs) – drugs like Prozac, Zoloft, Celexa, and Lexapro – which are the most commonly used antidepressants during pregnancy. However, much of the argument below would also apply to other antidepressants as well.
This post will review the evidence in this area and address the counterarguments.
This Isn’t Rocket Science
If you look at the 5 point argument above, it’s easy to see why the fact that these drugs aren’t safe in pregnancy is not rocket science. No one can argue with Point #1 above. Serotonin is a critical molecule for fetal development and normal function of the serotonin system is essential for proper human development. Serotonin plays a role in the formation of the brain, heart, lungs, gut, blood platelets, and other organ systems and we are discovering new and important functions for serotonin all the time. No one can argue (with a straight face) that serotonin is not crucial for human fetal development. Point 2 is also an established scientific fact. SSRI antidepressants are not naturally occurring substances. They are synthetic chemical compounds made in chemical factories and they disrupt the serotonin system. Scientists still aren’t exactly sure about all of the ways in which the SSRIs affect the human body, but it is clear that they do alter the serotonin system. Point 3 is also an established scientific fact. The SSRIs cross the placenta and enter into the fetus. Just putting Points 1, 2, and 3 together, one should easily conclude that the SSRI antidepressants are likely to cause some type of harm in fetal development. And this is precisely what we see in animal studies (Point 4) and human studies (Point 5). The animal studies show increased rates of heart problems, craniofacial abnormalities, death, brain abnormalities, and long-term neurobehavioral changes in the exposed offspring. Human studies show that SSRI use during pregnancy is associated with miscarriage, birth defects, preterm birth, preeclampsia, decreased fetal head size, newborn behavioral syndrome, seizures, neonatal EKG changes, childhood brain malformations, and long-term neurobehavioral issues like ADHD and autism. What do we think will happen when we send these chemicals into the developing baby throughout the pregnancy?
So Who Says Antidepressants are Safe in Pregnancy?
— Recognizing the tricks and traps of the Key Opinion Leaders
There are many “experts” from various centers out there emphasizing the safety of these drugs in pregnancy and they have been doing so over the past two decades. Much of the public does not realize that many of the leading centers of psychiatry and pregnancy (the so-called Perinatal Psychiatry units) and many of the “experts” have been (and are being) funded by the antidepressant makers themselves. The drug companies have been paying these experts (the so-called Key Opinion Leaders—or KOLs for short) and funding these units, and the message that gets delivered tends to downplay any safety concerns. This has been reviewed in detail here, here, and here.
Many of these experts seem to be following the same basic playbook that can be summarized as follows:
- Avoid discussing the basic science and animal studies: Why do we almost never hear the Key Opinion Leaders talking about the basic science data on the importance of serotonin in human development? Or how the SSRIs affect the serotonin system? Or the concerning findings from the animal studies? They virtually never comment on any of the first four points above. The reason they never go into this is that any discussion in this direction is bad news for their argument that the drugs are safe in pregnancy. The last thing these experts want is the public focusing on is how important serotonin is for fetal development and how the drugs disrupt that system and the concerning findings from the animal research.
- Emphasize how harmful depression is for pregnancy and how effective the antidepressants are at treating depression: Again and again, we hear from these experts (a) how depression leads to bad pregnancy outcomes and (b) how safe and effective the drugs are at treating depression. But, think about it: if (a) and (b) are correct then shouldn’t the research be showing improved pregnancy outcomes for the women on these drugs? Yet, this is never the case. In more than twenty-five years of study, we never see improved obstetrical outcomes in the group of women on these drugs. The SSRI-treated group consistently has worsened pregnancy outcomes (that is, more miscarriage, birth defects, preterm birth, newborn problems, and autism).
- Sow doubt in the public’s mind: This may be the oldest trick in the Key Opinion Leader playbook — dating back at least to the defense of cigarettes (and likely before then.) What the experts do here is to cast doubt on the available science and emphasize scientific uncertainty. Anyone involved in science knows that results are rarely completely certain — especially with complex studies in human populations. There is almost always some degree of scientific uncertainty whatever the issue. The KOLs capitalize on this. In the area of antidepressants and pregnancy one frequently hears the experts note that no randomized controlled trial has ever been done so we can’t really be sure if the harm we are seeing is due to the antidepressants or the underlying depression. This argument ignores the fact that when the human findings of harm (for example, persistent pulmonary hypertension of the newborn) exactly echo the animal studies (for example, rat studies showing that Prozac causes fetal pulmonary hypertension) then it is likely to be a drug effect. Also, randomized controlled trials aren’t always needed to determine harm (eg cigarettes). The science in this area is not perfect (and it almost never is in any area.) But there is now more than enough evidence of harm to conclude that these drugs are not “safe” in pregnancy.
- Argue that there may be harm but it is minor: Now that more and more evidence is emerging that antidepressants are linked to pregnancy harms, some KOLs have started to argue that the harms that we are seeing may be statistically significant but they aren’t clinically significant. This argument seems to make some sense at face value. For example, let’s say that you do a huge study with millions of people that shows that smiling all the time will increase your life expectancy — by 3 minutes! In this case, the 3 minutes of longer life may be statistically significant (especially with a study of millions of people) but it’s probably not clinically significant. So this argument can make sense in some arenas. But in the area of antidepressants and pregnancy complications, it does NOT apply. For example, with antidepressant use in pregnancy we see statistically significant increases in things like miscarriage, birth defects, and autism. By definition these things ARE clinically significant. If you are on an SSRI and have a miscarriage and lose that pregnancy, or have a baby with a heart defect, or a child with autism — that IS very clinically significant. The argument that the complications we see with antidepressant use in pregnancy are statistically significant but not clinically significant is utterly absurd. The second argument that the KOLs try to make in this area is that the absolute risks are low, but this is simply not the case. Some of the studies are showing very high risks. The recent study by Knickmeyer, et al showed that 18% of babies who are exposed to SSRIs in utero will have brain malformation. Toh, et al showed that 26.1% of women taking SSRIs will develop hypertension. Levinson-Castiel, et al demonstrated newborn behavioral syndrome in 30% of exposed newborns. These are not low risks.
- Accuse anyone trying to inform the public in this area of fearmongering: In September 2014 an article by Roni Caryn Rabin appeared in the New York Times on this topic. She, her editor, and the Times deserve congratulations for informing the public on this issue. Many of us who care for pregnant women and many of us who see the trend of increased use of antidepressants during pregnancy are very concerned about this. The best available scientific evidence — in over 25 years of study — is showing real harms with the use of these drugs in pregnancy. And the information on long-term problems in the children (eg autism, developmental delay) is truly frightening. We need more journalists and leaders in academic medicine bringing attention to this issue. Yet, those who speak out on this issue (or write about it) are often accused of fearmongering, being callous, or wanting pregnant women to kill themselves. These sorts of attacks are absurd. If we are going to “shout down” any journalist who dares to try to inform the public on this crucial issue, then we are going to end up with a very poorly informed public and potentially toxic chemical exposures in pregnancy.
- Compare antidepressants and depression to insulin and diabetes: This comparison is brought up all of the time by the KOLs and it just doesn’t hold water (i.e. it’s not scientifically accurate). When diabetic pregnant women get treated with insulin, they have improved blood sugars and better pregnancy outcomes. This has been demonstrated in numerous studies. The insulin-treated group has fewer miscarriages, fewer babies with birth defects, less hypertension, and better pregnancy outcomes. But when similar studies are done with depression, it’s the group on the antidepressants that consistently has the higher rates of complications. This is most likely because insulin is a natural protein that diabetics are deficient in. Giving insulin to a pregnant diabetic restores normal blood sugar levels and allows the fetus to develop normally (and the large insulin protein does not cross the placenta.) With depression, there is no “SSRI deficiency” that Prozac or Zoloft treats — that’s ridiculous. And, we now know, that there also appears to be no serotonin deficiency. That hypothesis is either dead or “last century thinking” depending on the source. Using antidepressants for depressed pregnant women is not like using insulin in diabetics.
- Portray concerned physicians, scientists, and journalists as bullies trying to tell pregnant women what to do: This is a particularly low-blow, especially when coming from KOLs who have been paid large sums of money over the years by the drug makers. No one should be telling pregnant women what to do in this area. The key issue is providing pregnant women and the public with full information, allowing them to make the best decisions for themselves, and supporting them in those decisions.
- Spin the research findings: Some might call this part of the KOL playbook “creative obfuscation.” A very good example of this comes from a recent op-ed piece in which the author wrote: “There are only a few studies that assess long-term outcomes among children whose mothers took antidepressants in pregnancy. Although the results were favorable, the research is rather lean.” This simply isn’t true. In many of the studies the results were NOT favorable. There are more than a few studies that have looked at longer-term outcomes and, in many, the children exposed to the antidepressants in utero are having significant problems: increased rates of autism, developmental delay, ADHD, motor problems, and structural brain abnormalities. To describe this body of research as a few studies with favorable outcomes goes beyond “spin.” This is just plain inaccurate.
The issue of the safety of antidepressants in pregnancy has never been more important as more and more women of childbearing age take these medications and exposure rates during pregnancy increase worldwide. The topic really is not that complicated. It’s just common sense that serotonin plays a crucial role in fetal development. The antidepressants cross the placenta, into the developing fetus, and chemically alter that serotonin system. So it shouldn’t surprise us that animal and human research is showing harm (and this is what we see clinically with actual patients as well.)
With good evidence that many patients will get as much benefit (or more) from non-drug approaches to depression such as psychotherapy and exercise, it is clear that these modalities should be emphasized in the treatment of pregnant women and women of childbearing age. No one should be telling pregnant women what to do, but the key issue is providing patients and the public with full information so that they can make the best decisions for themselves. They should be then be supported in their decisions and given the best care possible.
Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.
Mad in America has made some changes to the commenting process. You no longer need to login or create an account on our site to comment. The only information needed is your name, email and comment text. Comments made with an account prior to this change will remain visible on the site.