TCR: Dr. Hendrick, there has been a lot of confusing and seemingly contradictory data about the safety of SSRIs during pregnancy. What’s your take?
Dr. Hendrick: You’re right, it is confusing, but one reassuring and important point is that there is no evidence that the SSRIs or any other antidepressant are linked to an increased risk of congenital malformations. But what increasingly seems to be the case is that third trimester use of certain SSRIs is associated with an increased incidence of certain perinatal complications in newborns. For example, Lee Cohen at Mass General did a study comparing babies exposed to fluoxetine during the first trimester with babies exposed during the third trimester and found that the third trimester-exposed babies had more admissions to the NICU (neonatal intensive care unit), more jitteriness, respiratory problems, hypoglycemia and reduced APGAR scores. There’s also a program in Canada called The Motherisk Program and they saw the same phenomenon in children exposed to paroxetine in the third trimester. So the thinking now is going from worrying about first trimester exposure to worrying about third trimester exposure.
TCR: So does this imply that we should encourage women to discontinue antidepressants during the third trimester?
Dr. Hendrick: That’s controversial, because the perinatal complications reported are not sustained and they don’t seem to be linked with any long term sequelae. But if they could be avoided, so much the better, and one approach is to reduce the mother’s dose three or four weeks before her due date so that the baby is born with less exposure. This gives the mother less protection against depression as she’s about to deliver, but on the other hand there’s some evidence that if women immediately increase the dose on the day of delivery to the full dose, that this might be soon enough to forestall a relapse to depression in the post-partum period.
TCR: What do you estimate is the risk that a baby will have one of these perinatal complications?
Dr. Hendrick: In the studies that have reported a relationship between third trimester exposure and perinatal complications, the incidence of perinatal complications has been about 20-30%. In a recent study we found that 28 out of 138 babies exposed to SSRIs had these complications, and that there was a higher risk of low birth weight among babies exposed to high doses (60-80 mg) of fluoxetine. So one of our recommendations in this paper is for pregnant women to avoid high doses of fluoxetine. [American J of Ob/Gyn, 2003, 188:812- 815]
TCR: What about the risk of premature birth?
Dr. Hendrick: That’s another controversial issue. A recent paper by Simon and colleagues showed that children exposed to SSRIs were likely to be born about a week earlier on average than children in the control group. But as long as babies are not being born preterm (under 36 weeks), being born a week early does not pose a significant risk.
TCR: So your reading of the literature is that SSRIs do not confer a higher risk of frank preterm deliveries?
Dr. Hendrick: I think this is a question that needs to be explored further, given the conflicting data in the literature. Several studies have not found this risk, but others have. For example, a paper by Ericson and colleagues reported a slightly higher incidence of preterm birth. But this was a naturalistic study and there are many variables that were not controlled for, like the fact that these women might have been on other medications and might have comorbid conditions. Anxiety, for example, regardless of exposure to medication, has been linked very significantly with a higher incidence of preterm birth. One of the signals that initiates the process of labor in women is elevated CRH (cortcotrophin-releasing hormone), and if that hormone is elevated prematurely from stress it can fool the body into thinking that it should initiate the process of labor too early.
TCR: What is the risk of having a depressive relapse for a pregnant woman coming off medication?
Dr. Hendrick: Lee Cohen looked at women who had stopped antidepressants when they found out they were pregnant and he reported a 75% relapse rate over the course of the pregnancy. The risk of relapse is very substantial, and there’s certainly nothing about pregnancy that protects against relapse to depression.
TCR: What about breast-feeding and antidepressants?
Dr. Hendrick: Maternal depression has been shown to be associated with insecure attachment to the mother, lower self esteem later in life, greater incidence of aggressive behavior, and reduced scores on cognitive testing, among other problems. So we feel very strongly that women should be aggressively treated for depression during the post-partum. As far as breast-feeding goes, there is now good data from over 350 exposures showing that certain antidepressants can be used with no evidence of adverse effects on the child.
TCR: Which are the safest?
Dr. Hendrick: Our recommendations are that sertraline and paroxetine are the best choices, partly because we have so much data on them and also because the exposure to the infant is virtually always miniscule. In most cases, the blood levels show no detectable levels of medication even using very sensitive assays. Fluoxetine, on the other hand, seems to produce more exposure. It has a long half life, and an infant has only about a third of an adult’s ability to clear medications. In some cases, fluoxetine has been present at similar blood levels that we see in adults. There have been more reports of colic, jitteriness, and fussiness associated with fluoxetine in breast milk than any other antidepressant.
TCR: Much has been made of the timing of breast-feeding in relation to the time of taking an antidepressant. Is this something you talk to mothers about?
Dr. Hendrick: We do, and there is some data that sertraline levels peak about 6 to 10 hours post-dose, so that pumping and discarding breast milk at 6 to 10 hours post-dose will substantially reduce sertraline exposure to the baby.
TCR: Is this true for the other SSRI’s?
Dr. Hendrick: Our group has found the same time frame of peak levels for fluoxetine. However the peak is not as sharp as with sertraline, so pumping and discarding milk during that time frame does not get rid of as much medication as is the case for sertraline. For paroxetine, research by Stowe’s group at Emory has not shown a consistent relationship between time of dose and breast milk levels.
TCR: What about benzodiazepines and breastfeeding?
Dr. Hendrick: There’s not much data about this, which is too bad because it’s very common for patients to be anxious in the post-partum. Our recommendation is for women to use benzodiazepines that are shorter-acting and without active metabolites, so we use lorazepam (ativan) the most. Both klonopin and valium have been linked with isolated cases of respiratory depression and sedation in infants.
TCR: What about the use of benzodiazepines in pregnancy? We often hear about a potential risk of cleft palate.
Dr. Hendrick: The risk of cleft palate appears to be pretty small, at most rising from the 0.06% baseline rate to a rate of approximately 0.7%. This risk has been reported primarily with valium and xanax. But most meta-analyses have not shown an increased risk at all. So in general our recommendation is, just to be on the super-safe side, to try to avoid benzodiazepines during weeks 5 though 10 of gestation, when the lip and palate are forming. We also discourage benzodiazepine use late in the third trimester, because exposed neonates can be born sleepy.
TCR: Are there any medications for sleep that you regard to be quite safe during pregnancy?
Dr. Hendrick: Both Benadryl and low dose tricyclics, especially nortriptyline, appear safe.
TCR: On a different note, Dr. Hendrick, can you give our readers some advice on how to deliver our recommendations to pregnant women, especially from a medicolegal perspective.
Dr. Hendrick: Medicolegally, the concern seems to have diminished now compared to, say, five years ago, partly because data have accumulated and for the most part they are reassuring. What I recommend is that clinicians should follow some general guidelines for medicolegal considerations. First, if you don’t routinely work with pregnant women you should get a consultation and document this. Second, you should document both that you’ve gone over the risks of the medication, and that you’ve informed the woman that there is an overall 2-5% risk of babies being born with defects of some sort, that Mother Nature isn’t perfect. And third, there should be documentation that the mother is capable of understanding the information being provided to her because that can be an issue with working with mentally ill patients. Those are three key things. In addition, I’ll often make a note that I’ve provided the woman with papers that she can review at home.
TCR: What do recommend that we say when a patient asks us pointblank, “If it were your child, what would you do?”
Dr. Hendrick: If they ask me point blank, I’ll tell them what I would do. I try to go on the philosophy that I wouldn’t be recommending anything that I myself wouldn’t feel comfortable with. I also point out that if they were to see different clinicians they might get a very different point of view. But I would tell them what I would do, and I think this is one reason why women often want to come to a woman doctor who has had children of her own because there is a sense of greater empathy. Whether that’s true or not is debatable, but at least this is the perception among many women.