Article Text

Download PDFPDF

Antidepressant exposure before birth was not associated with language, physical, or behavioural development
  1. Guy M Goodwin, DPhil, FRCPsych, FRCP(Edin)
  1. University of Oxford Oxford, UK

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

OpenUrlCrossRefPubMedWeb of Science


To determine the cognitive and language behaviour of children who were exposed to tricyclic antidepressants or fluoxetine in utero.


Cohort analytic study.


Motherisk Clinic of the Hospital for Sick Children, Toronto, Canada.


Participants were considered to be mother and infant pairs. 80 women had taken tricyclic antidepressants during pregnancy; 40 during the first trimester, 36 throughout pregnancy, 2 during the first and second trimesters, and 2 during the first and third trimesters. 55 women had taken fluoxetine: 37 during the first trimester and 18 throughout the pregnancy. Pairs from these 2 groups were excluded if the drugs had been stopped before conception, ≥1 antidepressant had been taken, or exposure to known teratogens had occurred during pregnancy. 84 mother and infant pairs were included as controls. They had no known exposures that were recognised to harm the fetus and had had clinical appointments at approximately the same time as the mothers who received antidepressant medication.

Assessment of risk factors

Reported antidepressant use including duration of exposure during pregnancy. Other baseline risk factors were age at conception, gravidy, parity, abortion history, weight gain, socioeconomic status, intelligence quotient (IQ), alcohol use, smoking status, severity of depression, lifestyle, medical and nutrition status, history of sexually transmitted diseases, and type of delivery.

Main outcome measures

Children were assessed at 6 to 9 months and at 16 to 86 months of age. The second assessment included the Bayley Scales of Infant Development or the McCarthy Scales of Children's Abilities, Carey Temperament Scales, Achenbach Behavior Checklist, and Reynell Developmental Language Scales. Testing was done without knowledge of exposure to antidepressants before birth. Birth records were used to evaluate neonatal outcomes.

Main results

Multivariate analysis was done taking into account baseline variables. The 3 groups did not differ for gestational age at birth, birth weight, major abnormalities, perinatal complications, height and weight at testing, fronto-occipital circumference, or scores for all scales and indices (temperament, mood, arousability, activity level, distractibility, behavioural problems, or IQ).


Exposure before birth to tricyclic antidepressants or fluoxetine was not associated with differences in intelligence, growth, or language or behavioural development.


Depression is a common source of disability and is more commonly detected in young women than in any other group. Accordingly, efforts to increase the detection of depression and prolong the time for which drug treatment is offered will increase the probability that women will conceive while taking antidepressant agents.1

The findings of the study by Nulman et al support the view that serious perinatal problems are not seen at a higher rate in pregnant women prescribed either tricyclic drugs or fluoxetine compared with appropriately matched controls. This reassurance, while inevitably limited by the size of the sample, accords with the findings of previous studies.2 The present findings extend this reassurance by showing that the developmental assessment of children in their early years is unaffected by exposure to antidepressant drugs in utero.

The treatment of women who are pregnant, or more usually, those who have recently given birth and then become depressed should include the development of the child as an outcome variable as this study does. Impairment of the mother and child relationship in mothers who are depressed (and untreated) can lead to developmental delay and insecure attachment—at least in boys in socially deprived homes.3 Therefore, it may be best to treat pregnant women who are depressed not only from their personal perspective but also from the point of view of their unborn child. A study designed to answer this issue directly would be helpful. Depression after childbirth is also a common problem. A further implication of the present study may be that the babies of breast feeding mothers taking antidepressants are safer than is often assumed. A recent study has shown the efficacy of fluoxetine in postnatal depression.4

Whatever the evidence from statistical findings in large populations, it is highly likely that a woman who gives birth to a child with a birth defect may blame herself for having taken an antidepressant drug during pregnancy. Before prescribing therefore it is worth considering how a woman would react in such an eventuality. Furthermore, one would do well to remember that perinatal problems for mother or child, especially in older mothers, are distressingly common whether the mother is prescribed medicines or not.



  • Sources of funding: Motherisk Research Fund; Ciba Geigy Canada; Medical Research Council of Canada; Pharmaceutical Manufacturers Association of Canada.

  • For article reprint: Dr G Koren, Division of Clinical Pharmacology, the Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. Fax +1 416 813 7562.