select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to antidepressant use during pregnancy linked to autism

Publishing in the Journal of the American Medical Association a group of researchers have looked at the effect of the use of antidepressants during pregnancy and report that the use of selective serotonin reuptake inhibitors later in pregnancy can increase the risk of the child developing autism spectrum disorders.


Prof. Seena Fazel, Wellcome Trust Senior Research Fellow in Clinical Science and Professor of Forensic Psychiatry, University of Oxford, said:

“Whilst this is a seemingly sufficiently large study, there is an inherent limitation of this type of study design that can lead to a false positive association. The problem is that in this study all women who take antidepressants are categorised together despite the fact that women who take antidepressant medications in their second and third trimester are quite different from other women. Usually in these type of studies this problem, known as confounding, is dealt with by adjusting for background factors for which the researchers have added information. The problem with this specific study is that the authors don’t have the necessary information on potentially important background factors, which in this case will be genetic risks for autism.

“This type of problem is best exemplified in the large literature around a study of 600,000 infants that found associations between smoking in pregnancy and later adverse outcomes (including crime) – all of which turned out to be false positive associations when researchers in Sweden and elsewhere used more rigorous designs in follow-up studies. Here is one example but there are many:

“In summary, we cannot trust the conclusions around antidepressant use in pregnancy based on this type of study alone – more work will need to be done.”


Prof. Guy Goodwin, President of the European College of Neuropsychopharmacology (ECNP) and Professor of Psychiatry, University of Oxford, said:

“The design of this kind of study risks confounding by indication, which means the patients being prescribed a drug in later pregnancy may be more at risk of having an autistic child than patients who are not, for reasons we cannot identify but could be genetic, for example.

“The only fair test of the hypothesis is to use mothers as their own controls and look at the risk of autism in two different pregnancies in the same mother on and off medication. Such a study would not be impossible to do in very large databases, but the present study is far too small. To do a study where the mothers acts as their own controls is necessary to deal with unmeasured confounders, and such a study would need to be substantially larger than the current one.

“It is regrettable when inconclusive studies are misreported and cause distress to the many women who may need to take antidepressants to remain well during pregnancy.”


Prof. Ian Jones, Director, National Centre for Mental Health, Cardiff University, said:

This paper addresses an important issue for many women who face difficult decisions about starting or continuing antidepressant medication in pregnancy. The paper is consistent with some other studies published in recent years that have found an association between exposure to antidepressant medication in pregnancy and a higher risk of autism in children. However, the cause of this relationship remains uncertain. It is possible that the higher risk of autism spectrum disorder (ASD) is due to the medication, but it may also be due to the effects of the mood disorder for which the medication has been prescribed.

“Given that family studies show increased rates of mood disorder in the parents of children with autism it is also possible that an overlap in genetic factors between mood disorders and ASDs is also involved. It is estimated that approximately 3% of women take SSRIs in pregnancy. The baseline risk in the general population for ASDs in children is around 1%. The studies to date have suggested that the rate in children born to women with mood disorders taking SSRIS in pregnancy may be approximately double at around 2%. It is difficult to know whether this small increase in risk is due to the medication being taken, to the mood disorder itself, to an overlap in genetic vulnerability or to other factors associated with mood disorders and antidepressant medication.

“Depression in pregnancy and following childbirth (the postpartum period) is common, and is an illness with potentially serious consequences. It can be severe, with serious implications for the woman, her baby and her wider relationships. Suicide is a leading cause of maternal death in the UK. Women should not be taking medication in pregnancy if they don’t need to. For a woman taking an antidepressant and who is considering starting a family, or finds that she is pregnant, this is an excellent time to consider whether the medication is still required. Women should not stop their medication suddenly and, if they are concerned about continuing the antidepressant in pregnancy, they should discuss the options with their doctor. Antidepressants definitely have their place. Not only is suicide a major risk, but if a woman has an episode of depression it can have profound implications for the mother, baby and the whole family.

“There may be risks with taking antidepressants, as with other medications, but there are also significant risks from not receiving treatment and women who are concerned should discuss the balance of benefit and risk with their doctor.”


Prof. Louise Howard, Professor in Women’s Mental Health, King’s College London’s Institute of Psychiatry, Psychology & Neuroscience, said:

This study uses administrative registers to examine the possible association between antidepressant use and autistic spectrum disorders and its findings are inevitably difficult to interpret because of the study design. Associations found in observational studies cannot be assumed to be causal as the associations could be due to other factors that have not been accounted for in the analysis i.e. confounding factors.

“The authors here try to address the confounding effect of depression which may itself be causally related to autism but they can only adjust for a history of depression due to the nature of their source of data; however, if depression does confound the association between antidepressants and autism, data on the severity of depression during pregnancy would need to be included in their model. The finding that use of two antidepressants was more likely to be associated with autism suggests that indeed severe depression (for which two antidepressants may be prescribed) may be the risk factor here. There are also other risk factors that the authors acknowledge may confound the association that could not be included in their analysis such as smoking.

“The study does not find an association between some types of antidepressants whereas it confirms previous findings of an association with SSRIs. This may be due to the smaller numbers of antidepressant prescriptions for non SSRI prescriptions and therefore reflect a lack of statistical power to detect a significant association. The direction of effect is not dissimilar to that for SSRIs and again may reflect severity of depression or associated risk factors which may be associated with autism rather than the antidepressant.

“Clinicians and women are therefore still left with considerable uncertainty regarding the possible risk of autism from antidepressants. In practice it is preferable to use psychological treatments for depression in pregnancy where possible in order to avoid in utero exposure to medication. However where depression is severe and unlikely to respond to psychological interventions alone, then the risks of untreated depression (both to mother and the fetus as highlighted in last week’s confidential enquiry into maternal deaths and the Chief Medical Officer’s report into women’s health) will mean that the benefits of treatment with medication will outweigh the possible risks.”


Prof. Emily Simonoff, Professor of Child and Adolescent Psychiatry, King’s College London’s Institute of Psychiatry, Psychology & Neuroscience, said:

“The article by Boukhris et al in the present issue of JAMA Pediatrics points to an association between the use of antidepressant medication during pregnancy and the subsequent diagnosis of an autism spectrum disorder (ASD) in the child. The authors use data from a Canadian registry to explore a possible link between prenatal antidepressant exposure in pregnancy and ASD diagnosis in the children up to the age of 11 years. They find that the association is only significant when SSRIs are considered, on their own or in combination (although this is the only class of antidepressants used with sufficient frequency to identify a small to moderate effect), and also that it appears to be specific to their use in the second or third trimester of pregnancy.

“In recent years a number of observational studies have explored the possible link between prenatal antidepressant use and subsequent neurodevelopmental disorder in children, with varying results. A recent systematic review and meta-analysis of the previously available studies reported a significant association between prenatal SSRI exposure and subsequent ASD diagnosis, although some individual studies do not find this link (Man et al., 2015).

“Observational studies such as the present one can often identify associations but even when possible confounders are considered, a causal link cannot be inferred. One a well-recognized link is between ASD and increased risk of affective disorder in family members.  As pointed out in an accompanying editorial by King, there is evidence for a genetic overlap between the disorders. This raises the possibility that some women who experience depression may also be at increased risk of having a child with ASD, and that use of SSRIs during pregnancy is indexing this risk, rather than causing ASD in children.

“Further research using additional datasets and other methods is required. One important observational method examines sequential pregnancies in women where there is discordance in prenatal exposures, to determine whether the link is with the exposure or other factors associated with the mother. Animal studies are also required to better understand the possible pathophysiology.

“As pointed out by the authors and accompanying editorial, the increased rate of ASD associated with SSRI use is relatively small and there are potentially many adverse effects of discontinuing antidepressants amongst women at risk of recurrent affective disorder who may experience increased stress as a direct result. The current findings should not lead to a blanket change in policy. All pregnant women, as well as those planning a pregnancy, who are taking SSRIs should discuss the therapeutic options with their doctor, bearing in mind that the present findings do not necessarily indicate a causal role of the medication.”

Man, K. K., Tong, H. H., Wong, L. Y., Chan, E. W., Simonoff, E., & Wong, I. C. (2015). Exposure to selective serotonin reuptake inhibitors during pregnancy and risk of autism spectrum disorder in children: A systematic review and meta-analysis of observational studies. Neuroscience and Biobehavioral Reviews, 49c, 82-89. doi: 10.1016/j.neubiorev.2014.11.020


Antidepressant Use During Pregnancy and the Risk Of Autism Spectrum Disorder in Children’ by Boukhris et al. published in JAMA Pediatrics on Monday 14th December. 


Declared interests

Prof. Fazel: “I once received travelling expenses by Janssen (that don’t make antidepressants) to attend a conference.”

Prof. Goodwin: “I have advised companies on the development of drugs to treat depression.”

Prof. Howard: “I was lead author of the chapter on perinatal mental health in the CMO report 2014 on women’s health.”

Others: None declared

in this section

filter RoundUps by year

search by tag