TCPR: Dr. Hendrick, earlier this decade, most experts felt that SSRIs were relatively safe in pregnancy, but lately the pendulum seems to be swinging in the other direction. What do you consider to be the major risks of SSRIs in pregnancy?
Dr. Hendrick: The adverse outcomes can be broken into four general categories: first, a greater risk of perinatal complications in the newborn right at birth; second, birth defects, especially cardiovascular anomalies like atrial septal defects; third, persistent pulmonary hypertension of the newborn; and finally, miscarriage. But it is important to realize that while SSRIs have some adverse effects, untreated depression during pregnancy has been found to have higher rates of some of these outcomes as well.
TCPR: Can you describe some of this research?
Dr. Hendrick: One good study, with nearly 120,000 participants (Oberlander et al., Arch Gen Psychiatry 2006;63:898-906), compared three groups of pregnant women born in British Columbia: depressed women who took SSRIs, depressed women who did not take SSRIs, and a control group of non-depressed women who were not on medication. What they found was that depressed women who were not on medication had worse outcomes than non-depressed women (also on no medication). Specifically, their babies were more likely to be preterm, have low birth weight, and have respiratory distress. Nevertheless, the greatest risk of poor outcome was in the SSRI-exposed moms. So it looks like depression alone can cause poor outcomes, and that SSRIs seem to cause independent risk.
TCPR: And how might untreated depression lead to these poor outcomes?
Dr. Hendrick: Untreated depression has been linked to missed prenatal doctor’s visits, substance abuse, and smoking. In general, when studies have controlled for maternal health habits, depression alone is not linked to poor outcome; it is the poor health habits of depressed women that seems to be crucial.
TCPR: That’s somewhat reassuring. In my own practice I commonly see pregnant women who are depressed and yet do not want to take medication. It sounds like what you are saying is that as long as they appear to be taking care of themselves and their health, I needn’t be concerned that the depression per se will harm the fetus.
Dr. Hendrick: Yes, that’s right. As long as they are gaining weight adequately and following up with all their prenatal visits, just the fact that they are depressed shouldn’t lead to a worse outcome. However, if they have depression with anxiety as well, which is very common in pregnant women, then there are some theoretical concerns, because higher levels of stress hormones early in the third trimester has been linked with preterm labor. So, if a woman is highly anxious, the anxiety itself, even with good health habits, could lead to a greater risk of an adverse outcome.
TCPR: Let’s talk about Paxil (paroxetine). Most authorities consider this medication to be relatively contraindicated in pregnancy because of the risk of cardiac abnormalities. But the studies have often clashed with one another. What’s your take on the issue?
Dr. Hendrick: The studies can be confusing. One of the general rules of thumb I use in interpreting pregnancy studies is to look at how consistent the data are. If only a single study finds a certain outcome, then it is not going to be as compelling as if several studies have found the same poor outcome. Likewise, the “poor outcome” should be the same birth defect. In this case, there have been a variety of cardiovascular defects linked with SSRIs, and that argues a bit against the link between a prenatal exposure and the cardiovascular defect.
TCPR: Have the Paxil studies consistently shown any particular cardiovascular defect?
Dr. Hendrick: Yes, a number of studies have linked paroxetine with atrial septal defects and ventricular septal defects, and this concerns me enough that I recommend that paroxetine not be used in pregnancy. In addition, paroxetine is the only SSRI that has been bumped down to category “D” by the FDA (there is evidence of risk in humans), making it a medical-legal risk. But it’s important to note that not all experts agree with the negative perspective on Paxil, and seemingly contradictory results are published frequently. That said, I don’t worry too much about the cardiovascular defects of SSRIs, because even if there is a genuine association, the risk appears to be very tiny. My biggest concern is with the respiratory distress that has been linked with SSRI exposure.
TCPR: What is the risk?
Dr. Hendrick: There have been multiple reports of perinatal complications occurring in about 15%-30% of newborns exposed to SSRIs and SNRIs near the end of pregnancy. One of the most consistent of these complications is respiratory distress. Other complications include hypoglycemia, tremulousness and lower Apgar scores. These problems appear short-lived, resolving within 4-5 days. To my knowledge there have been no lasting sequelae or serious complications. Nevertheless, we prefer to minimize the risk of these complications whenever possible.
TCPR: What do you advise women to do to avoid these risks?
Dr. Hendrick: As patients get near term, as long as they are doing pretty well, I will try to reduce the dose of an SSRI so that the baby is born with the least amount of medication on board. This increases the risk of post-partum depression, but you can go right back to the original dose as soon as the baby is born. A lot of my colleagues who work with pregnant women don’t like to do this. But in the end, one of the points that I emphasize to trainees is that it is not our decision to make; we can only present the risks and benefits to the mother and her partner and let them choose the risk they feel most comfortable with. We can just present the data and let the patient decide.
TCPR: What other outcomes have been associated with SSRI use?
Dr. Hendrick: There have been reports of a higher rate of miscarriages in women taking SSRIs. It is very hard to tease out whether this is due to medication exposure or other factors. In general, women who take antidepressants are more likely to be older and to smoke, and these are two risk factors for miscarriage. If I have a patient who is trying to get pregnant and who already has risk factors for miscarriage, I would recommend that she not be on SSRI while she is trying to conceive.
TCPR: What are the overall risk factors for miscarriage?
Dr. Hendrick: The main ones are being older than 35, being a smoker, having a history of a previous miscarriage, having delivered a baby recently, or having a chronic disease like diabetes or lupus.
TCPR: You mentioned persistent pulmonary hypertension. What is this, and how strongly is it associated with SSRIs?
Dr. Hendrick: At birth, the newborn must use its lungs for the first time, and the arteries bringing blood to the lungs have to relax and dilate in order to bring enough blood flow through the lungs to oxygenate the blood. In persistent pulmonary hypertension of the newborn (PPHN), the arteries do not relax enough, and the baby has trouble getting enough oxygen. Sequelae can range from minor tachypnea to really severe cases where the baby can die or have brain damage. In 2006 a study was published in The New England Journal of Medicine that found a significantly higher rate of exposure to SSRIs in newborns with PPHN. However, in my opinion, this paper was unnecessarily alarmist because it gave the impression that children exposed to SSRIs could have serious adverse outcomes, when in fact the cases of PPHN were all mild, and there were no long-term problems in this particular group.
TCPR: Is there any particular antidepressant that you favor in pregnancy?
Dr. Hendrick: The one I use the most is Prozac (fluoxetine), because it has the most studies, but it is a close call between Prozac, Zoloft and Celexa. I look at other things that carry more weight, like whether the woman is already on a medication that is working well for her. If she is already on Celexa, we will just keep her on Celexa, I wouldn’t switch. I would prescribe the lowest possible dose.
TCPR: Is there any research documenting an effect of dose on adverse outcomes?
Dr. Hendrick: Yes, a colleague at UCLA found a direct inverse association between dose and risk of preterm birth (Suri R, et al., Am J Psychiatry 2007 Aug;164(8):1206-13). She found that the women taking a higher dose of an SSRI were about twice as likely to have a preterm birth than the women on a lower dose.
TCPR: What is your approach to discussing the various risks with patients?
Dr. Hendrick: We always inform (and document that we informed) pregnant women that nature is not perfect: there is an underlying 2%-4% risk that a baby will have a birth defect. We try to keep women from expecting that every baby is going to be perfect. It is also important to emphasize that the data are still fairly limited, and there are all sorts of things that we don’t yet know.