Intended for healthcare professionals

CCBYNC Open access

Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: population based case-control study

BMJ 2013; 346 doi: (Published 19 April 2013) Cite this as: BMJ 2013;346:f2059

Re: Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: population based case-control study

Maternal antidepressant use and autistic spectrum disorder: Fetal reprogramming or pre-conception epigenetic phenomenon, conferring what real-world magnitude of risk?.

Alison .M. Bratt BSc. (Hons), PhD: Lecturer in Pharmacology, Medway School of Pharmacy, Universities of Kent & Greenwich at Medway, Central Avenue, Chatham Maritime, Kent, ME4 4TB.

Comment on the article: Rai D, Lee B K, Dalman C, Golding J, Lewis G, Magnusson C. (2013) Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: Population based case-control study. B.M.J. 346: 1-15.

Rai et al, (2013) very recently published results of their research study testing the hypothesis that antidepressant drug use during pregnancy was associated with increased risk of autistic spectrum disorder (ASD) in the offspring. This large case controlled study conducted in Stockholm, Sweden between 2001-2007, found that in utero exposure to both SSRI’s and monoamine oxidase inhibitors was associated with a two-fold risk for birth of a child with ASD. These findings corroborate earlier independent results collected by Croen at al., (2011) in the USA which found any exposure to SSRI’s while pregnant gave a 2.2-fold increase in risk of ASD diagnosis, with first trimester exposure raising the risk 3.8-fold.

Preclinical research has demonstrated that homeostatic serotonergic tonus within the central nervous system is necessary for normal brain morphogenesis in utero (Oberlander, 2012). Indeed exposure of pregnant rats to clinically relevant doses of antidepressant drugs (ADD’s) disturbs both the architecture and functioning of the young rat brain causing aberrant behaviours, (Darling et al., 2011; Homberg, 2009; Simpson et al., 2011), which have some degree of translational relevance to the clinically diverse presentation of autism. The two major brain circuits affected by prenatal ADD exposure in rats appear to be the somatosensory cortex and dorsal raphe nucleus (DRN)- medial prefrontal cortex (mPFC)-amygdaloid corticolimbic circuit, underpinning sensory appreciation and stress-mediated behaviours respectively, (Darling, 2011)

Both the Rai and Croen studies could be criticised for being too small and inadequately powered, by including too few affected individuals, and clearly it is necessary to amass data from a larger population to strengthen and further these findings. One possible way to achieve this might be to utilise such national databases as the Autism Spectrum Database U.K. funded by Autistica. This growing database currently has approximately 700 families registered and is growing rapidly throughout the U.K. Data could be extracted from this database regarding previous incidents of depressive episodes and psychopharmacological treatment exposure from both parents of children with ASD, allowing a comprehensive capture of a variety of possible biological, experiential and contextual confounding variables.

There are several noteworthy points of interest that require further investigation arising from the study by Rao et al., 2013. Guidelines for clinical practice both in terms of informing prescribers and in final prescribing practice of ADD’s in pregnancy should be considered. Given the apparent ambiguity in the literature regarding any superior efficacy of SSRI’s in comparison with some non-drug therapies, (Khan et al., 2012), it would be prudent to advise treating pregnant women with ADD’s only when non-drug interventions fail, and when symptoms are sufficiently severe. Untreated depression during pregnancy of course carries its own risks, for both the mother and the foetus, but these should be rigorously monitored and treated with individualised intensive therapies on a case by case basis.

At present the British National Formulary (BNF) gives the following guidance for the prescription of SSRI’s in pregnancy, that “Manufacturers advise against use in pregnancy unless the potential benefit outweighs the risk. There is a small increased risk of congenital heart defects when SSRI’s are taken during early pregnancy. When SSRI’s are used during the third trimester there is a risk of neonatal withdrawal symptoms”. It might be prudent, especially in light of growing evidence, to consider an amendment to BNF wording to inform of an association between antidepressant use in pregnancy and possible resultant ASD.

It would be interesting to consider the possibility that the children associated with prenatal ADD exposure may constitute a phenotypically distinct subgroup of children with ASD? The Rai et al, 2013 paper showed that these children did not present with intellectual impairment, and were considered more “high functioning” on the autistic spectrum, mirroring characteristics of the group of autistic children which latest demographic data on diagnostic trends have shown to be increasing in number throughout the USA (Autism & Developmental Disabilities Monitoring Network 2012) and worldwide. Whether these children experience specific clusters of symptoms, such as more affective (high anxiety or irritability) or sensory dysregulations would have implications for how best to support them in home and educational settings.

Controversially, consideration should be given to whether exposure to SSRI’s during a “pre-conception” period confers any additional risk for ASD? The serotonin specific transporter SLC6A4 has been found to be present in mouse oocytes and appears to play a role in follicular growth and meiotic maturation, (Amireault & Dube, 2005). This raises questions surrounding whether serotonin guides the same reproductive processes via the same mechanisms in man and if SSRI exposure could potentially damage early embryogenesis?

Considering the challenges faced by individuals and their families living with autism, whatever the real-world risk maternal antidepressant use carries, it is a risk worth understanding.

Amireault P, Dube F. (2005) Serotonin and its antidepressant-sensitive transport in mouse cumulus-oocyte complexes and early embryos. Biol. Reproduct. 73: 358-365.
Autism Spectrum Database U.K. :
Croen L A Grether J K, Yoshida C K, Odouli R, Hendrick V. (2011) Antidepressant use during pregnancy and childhood autism spectrum disorders. Arch. Gen. Psychiatr. 68(11): 1104-1112.
Darling R D, Alzghoul L, Zhang J, Khatri N, Paul I A, Simpson K L, Lin R C S. (2011) Perinatal citalopram exposure selectively increases locus coeruleus circuit function in male rats. J. neurosci. 31(46): 16709-16715.
Homberg J R, Schubert D, Gaspar P. (2009) New perspectives on the neurodevelopmental effects of SSRI’s. Trends in Pharmacol Sci. 31(2): 60-65.
Khan A, Faucett J, Lichtenberg P, Kirsch I, Brown W A. (2012) A systematic review of comparative efficacy of treatments and controls for depression. PLoS ONE 7(7):
Oberlander T F. (2012) Fetal serotonin signalling: Setting pathways for early childhood development and behaviour. J. Adolescent Health 51: S9-S16.
Prevalence of autism spectrum disorders-Autism and Developmental Disabilities Network, 14 sites, United States, 2008. MMWR Surveill Summ (2012) 61: 1-19
Simpson K L, Weaver K J, de Villiers-Sidani E, Lu J Y F, Cai Z, Pang Y, Rodriguez-Porcel F, Paul I A, Merzenich M, Lin R C S. (2011) Perinatal antidepressant exposure alters cortical nerwork function in rodents. P.N.A.S. 108(45): 18465-18470.

Competing interests: No competing interests

01 May 2013
Alison M Bratt
Lecturer in Pharmacology
Medway School of Pharmacy
Universities of Kent & Greenwich at Medway, Central Avenue, Chatham Maritime, Kent, ME4 4TB