 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
I am pleased that the letter by Norton et al gives us the opportunity to clarify the significance and limitations of the findings of our study on the association between maternal use of selective serotonin reuptake inhibitors (SSRIs) during pregnancy and neonatal neurobehavior. As we described in our article, maternal depression during pregnancy is an important public health concern, with implications not only for the health of the mother but also for the health and development of the infant. Treatment of this condition has focused on mothers using SSRIs, with the assumption that these medications are safe for the developing fetus. Our concern is that this assumption is based on research that indicates that prenatal SSRI exposure has little or no impact on development, because exposed infants are born mostly of full birth weight, at full term, and show few physical anomalies. However, over the past 30 years, significant research literature has demonstrated that, in addition to these measures, neurobehavioral assessments provide a measure of a dimension of development that may or may not occur with problems in physical growth.1,2 The question is whether prenatal SSRI exposure may be associated with neurobehavioral perturbations among infants who do not show physical signs of atypical development.
To examine this question, we replicated the basic design used in previous studies in which few or no anthropometric effects of prenatal SSRI exposure were found. That is, we compared infants with prenatal SSRI exposure to those infants without SSRI exposure but added the use of neurobehavioral measures to assess infant development. We found that SSRI-exposed infants, although they were full birth weight, born at term, and showed no physical anomalies, demonstrated a wide range of neurobehavioral characteristics that are typical of infants with problematic prenatal environments.
The first concern of Norton et al is that our study lacked an additional control group that consisted of infants of untreated, depressed mothers. We directly addressed in the article the importance of this and other methodologic issues for future research. We agree that adding this comparison group would help clarify whether differences in neurobehavioral outcomes were the result of maternal depression, prenatal SSRI exposure, or a combination of both. There are also several other considerations that would benefit future research, including an analysis of the severity of maternal depression, the type of SSRI and its dose level, and the timing during pregnancy in which the medication was used. Random assignment to treatment groups would be necessary also to determine causality of effects. Although all of these methodologic considerations would benefit future research, they are unnecessary to establish the point that measures of neurobehavioral development are sensitive to group differences between SSRI-exposed and nonexposed infants who are otherwise full birth weight, born at term, and show no physical anomalies.
Their second concern regards the use of 1-tailed tests in our statistical analyses. Consistent with standard and correct practice, we used 1-tailed tests to compare groups because we had specific directional hypotheses. In the review of the literature and rationale for each outcome measure, we described how infants with a wide range of adverse prenatal conditions, including prenatal SSRI and other drug exposure, typically have more tremors and startles, greater motor activity, a more restricted range of states, increased rapid-eye-movement activity, and fewer numbers of rhythms in heart rate variability. The directions of some of these effects were specific to those suggested by case studies of prenatal SSRI exposure. Perhaps we could have been more explicit about the hypotheses of the study, but we thought our review and rationale were clear on what we expected and did find.
Their third concern regards the risk to the mother of discontinuing the use of antidepressants during pregnancy. As we discussed in our article, maternal depression, in itself, is a danger to both the mother and the infant. We are clearly not advocating the cessation of using SSRIs to treat maternal depression when it is clinically necessary. Additionally, our research does not address under which conditions SSRIs should or should not be used. However, as we stated in the article, we believe any effects on the infant attributable to prenatal SSRI exposure should be entered into a careful cost-benefit discussion between the physician and patient. Contributing to the discussion of this cost-benefit equation is literature replete with evidence of the benefits of SSRI treatment and the problem of depression, including our own research showing how women's depression may negatively affect their responses to infants' social signals.3 Our concern continues to be, however, that research regarding the costs of such treatment has suffered from the conclusion that, in the absence of physical signs, there may be no detrimental impact on infant development of prenatal SSRI exposure. Our study shows that neurobehavioral indices are sensitive to the conditions that differentiate SSRI-exposed and nonexposed groups in the absence of atypical physical signs. We believe that the addition of neurobehavioral measures in future research may contribute meaningfully to the cost-benefit analysis in physician-patient discussions.
REFERENCES
Related articles in Pediatrics:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2004 American Academy of Pediatrics. All rights reserved. |
||
|
||
